(ScareCrow Intelligence team: Cevdet Bulut, Joshua Brett, Nikolay Litov)

Artificial intelligence (AI) is increasingly developing within a global technological landscape dominated by a US–China duopoly. From foundation models to cloud computing infrastructure, many of the technologies underpinning the AI revolution remain concentrated in the hands of a small number of firms, raising concerns about data sovereignty, unequal access to innovation, and the uneven distribution of economic benefits.

We had the pleasure of hosting the ScareCrow Intelligence team at the University of Oxford to discuss the role of artificial intelligence in the food and agricultural sector. The founders argue that the future of AI should not be understood solely through the lens of geopolitical competition. ScareCrow is a UK-based company headquartered in Kent that develops automated, turnkey solutions for early warning systems and proactive risk management in agriculture. Their experience demonstrates that, when applied to local challenges, AI can serve as a powerful instrument for economic development—particularly in low- and middle-income countries.

Beyond the Duopoly: AI as an Instrument for Global Development

Agriculture offers one of the clearest examples of this potential. Advances in AI-powered image recognition, remote sensing, drones, and Internet of Things (IoT) sensors now make it possible to produce high-resolution maps that identify crop stress, pest infestations, disease outbreaks, and variations in soil and water conditions. These technologies allow farmers to apply fertilisers, pesticides, and irrigation selectively rather than uniformly across entire fields, improving both productivity and environmental sustainability. Precision agriculture of this kind not only reduces input costs but also limits the excessive use of chemicals, making production more resource-efficient.

Equally significant is AI’s capacity to power early-warning systems for crop disease. By combining environmental data including humidity, temperature, rainfall, wind patterns with satellite imagery, drone photography, and plant images captured in the field, machine learning models can flag disease risk before symptoms are visible to the naked eye. Instead of following routine spraying schedules, producers can intervene only when predictive models indicate a genuine threat. This is the shift from reactive to predictive agriculture: fewer unnecessary pesticide applications, better yields, and lower production costs.

One of the most compelling parts of the seminar was ScareCrow’s own trajectory in Ghana. The company illustrates how technological innovation can be adapted to the realities faced by smallholder farmers, who often operate under significant constraints in terms of finance, infrastructure, and access to agricultural extension services. ScareCrow’s work shows that AI in Africa has potential to respond to concrete developmental challenges. Precision mapping of farmland, identification of vulnerable areas, early disease detection, and data-driven decision support all help farmers increase productivity, use fewer inputs, and build resilience to climate-related risk. Rather than replacing agricultural expertise, AI augments farmers’ capacity to make informed decisions under conditions of uncertainty.

This view aligns closely with the position of the Food and Agriculture Organization of the United Nations (FAO), which identifies digital agriculture and AI as important tools for strengthening food security, productivity, and climate resilience in developing economies. The developmental significance of AI, in other words, extends well beyond the technology itself — its greatest contribution may lie in helping countries address long-standing structural challenges in agriculture, where productivity gains carry direct implications for poverty reduction, food security, and inclusive growth.

The seminar ultimately challenged a common narrative surrounding artificial intelligence. While contemporary discussions often emphasise the strategic rivalry between the United States and China, the experience presented by ScareCrow suggests that a complementary conversation deserves greater attention: how AI can be harnessed by entrepreneurs to solve context-specific problems in developing countries. The future of AI should therefore be evaluated not only by the sophistication of its models, but also by its capacity to improve livelihoods, strengthen food systems, and contribute to sustainable development where these technologies are needed most.

Beyond ScareCrow’s own trajectory, one of the greatest opportunities for AI in agriculture lies in its ability to bring together information that has traditionally existed in isolation. Farmers typically draw on weather forecasts, field observations, historical records, satellite imagery, and agronomist advice — but these sources are rarely connected in a way that supports quick decision-making. ScareCrow’s Predictive Intelligence Hub analyses these different streams simultaneously, helping farmers spot patterns that would otherwise be difficult to predict. The result isn’t simply more data; it’s more timely, actionable insight. The FAO describes digital agriculture as a key enabler of more productive, resilient, and sustainable food systems, particularly as farmers face growing pressure from climate change and resource constraints.

AI’s role doesn’t stop at the farm gate. Producing more food won’t solve global food insecurity if a large share of it is lost before reaching consumers. AI-powered forecasting can improve storage, transport, and distribution by predicting demand, flagging potential disruptions, and reducing waste across agricultural supply chains, helping businesses and governments decide where food is needed most, while reducing environmental impact. According to the FAO, digital technologies are becoming just as important for strengthening entire agrifood systems as they are for production itself.

Ultimately, though, the success of AI in agriculture shouldn’t be measured by the sophistication of its algorithms alone, but by whether it helps farmers make better decisions. The World Bank notes that AI delivers the greatest value when it turns complex information into practical advice farmers can trust and act on, namely complementing local knowledge and experience rather than replacing it. As digital technologies become more accessible, locally developed innovations have real potential to help farmers produce more with fewer resources, strengthening food security while supporting more sustainable agricultural systems worldwide.

The experience shared by ScareCrow Intelligence at Oxford this week is far from an isolated case. Research from the Brookings Institution highlights how digital agronomy tools developed with CGIAR have delivered measurable gains for farmers in Ethiopia, Nigeria, and Mali by offering location-specific fertiliser and planting guidance rather than one-size-fits-all recommendations. World Bank analysis points to the same trend: the region’s agri-food tech sector has grown rapidly in recent years, with private investment expanding as platforms combining machine learning, weather forecasting, and low-connectivity communication tools reach growing numbers of farmers. What unites these initiatives is a shift away from generic, imported AI solutions toward tools built around the everyday realities of agriculture.

Seminars like this one are a reminder that meaningful AI innovation isn’t confined to the handful of firms dominating foundation models and cloud infrastructure. It’s worth balancing this optimism with an honest look at AI’s limitations in agricultural development. The World Bank cautions that AI in agriculture carries real risks of exclusion and bias, particularly where systems are designed around farmers who already have smartphones, reliable connectivity, and stronger digital literacy, potentially widening the gap for those without these advantages. Questions of data governance, farmer control over their own information, and the environmental cost of expanding digital infrastructure also remain unresolved. As the World Bank puts it, AI is not a universal solution to agricultural challenges; its usefulness depends on good data, reliable connectivity, and institutions farmers actually trust.

From precision agriculture and early disease detection to the broader question of who AI is actually built for, these threads reflect a conversation that is only becoming more relevant as digital tools continue to reshape agriculture across the world. As more entrepreneurs take up this work, the question is no longer whether AI has a role to play in global food security — but how quickly, and how equitably, that role can be scaled.

NATO’s $40 billion Drone Edge is three admissions at once: that cheap, smart machines have ended the West’s monopoly on precision, that power inside the alliance now flows to whoever adapts fastest rather than spends most, and that the decision to kill is quietly passing from the operator to the algorithm.

In early 2026, the world’s most expensive military did something telling: it copied its enemy’s cheapest weapon. The American LUCAS drone — roughly $35,000 apiece, reverse-engineered from Iran’s Shahed — went into combat in the Middle East within days of its debut. Months later, on 7 July, NATO’s leaders gathered in Ankara and launched Drone Edge: more than $40 billion for counter-drone capabilities over five years, and five times as many drone operators by the end of 2027. The alliance’s official text states, flatly, that drones have become “a decisive factor on the battlefield.” It is a procurement programme presented as routine. It is better read as a confession — three of them, folded into one press release. And what changed the battlefield was not the drone. It was the software inside it.

THE END OF A MONOPOLY

For most of history, military power meant numbers. Since the Cold War it meant precision — hitting exactly what you aimed at, which only rich states could afford. Cheap drones, cheap sensors and, above all, cheap guidance software have now fused the two. Scholars Michael Horowitz and Lauren Kahn call the result “precise mass”: accuracy in bulk, available to anyone. Eric Schmidt, who has made repeated trips to Ukraine over the last four years, puts it more sharply. The fighting, he argues, “is no longer a war of platforms. It’s a war of systems” — the decisive unit is not the drone but the architecture that lets a force see, decide and strike faster than its enemy. The hardware is almost incidental; the edge lives in code. And code does not respect a defence budget.

The proof came from three theatres. In Ukraine, small attack drones now cause roughly 60–70 per cent of frontline casualties on both sides, and output climbed from about 800,000 in 2023 toward four million in 2025. In June 2025, Operation Spiderweb smuggled 117 drones costing $600–1,000 each deep into Russia and destroyed strategic bombers worth billions — at least ten of them, by satellite count. In the Middle East, Iran fired Shahed-type drones at American bases, Israel and the Gulf and exposed a brutal piece of accounting: killing a $20,000–50,000 drone with a Patriot missile costing over $4 million hands the attacker a ratio of eighty or two hundred to one. Hence LUCAS: if you cannot afford to shoot cheap drones down, you learn to fire them back. Then it came home — nineteen Russian drones crossed into Poland in September 2025, and unidentified ones shut Copenhagen’s airport for hours. The humiliation was not that NATO could not stop them. It was that it could not afford to. The West’s monopoly on precision was finished.

THE NEW HIERARCHY

Finished monopolies redistribute power — the second confession. If the edge lives in software rather than steel, the states that rise are the fastest to reorganise around code, not the richest; and that capacity is not distributed the way GDP is. Hence the summit’s unfamiliar cast. Turkey — the host, and no accident that the alliance’s centre of gravity now tilts toward its southern and eastern flank — fielded Baykar, which by its own government’s account exported $2.2 billion in 2025 and remained the world’s largest armed-drone exporter for a third year.^[3] Ukraine, not even a member, has become the alliance’s indispensable supplier of the one thing no factory produces — combat-tested know-how — and allies now treat it as a security provider, not a supplicant. Tiny Latvia co-leads a twenty-nation drone coalition with Britain.

This is not a shift away from America, whose technological weight is unmatched and which is racing the same way, aiming to buy 200,000 FPV drones of its own by 2027. It is a shift away from the slow — and Europe’s legacy giants are slow. Mario Draghi’s 2024 report found that 78 per cent of some €75 billion in EU procurement went to suppliers outside the Union, while the continent fields a dozen types of main battle tank to America’s one.^[5] The mindset is slower still: in March 2026 the chief executive of Germany’s Rheinmetall belittled Ukraine’s drone innovations — the authentic voice of an industry that prizes costly, exquisite platforms and mistrusts the cheap and the fast. Drone Edge is the first alliance programme to admit that its hierarchy now tracks adaptation, not wealth.

THE THIRD CONFESSION

The slow are punished; the fast automate. That is the third confession, and it is the one Drone Edge funds without daring to name. Look first at how desperately both sides are trying to keep a human in control. A drone’s radio link is its weakest point — jam it and the weapon drops — so Russia and Ukraine now fly drones on spools of hair-thin fibre-optic cable, physically tethered, impossible to jam. The counter is almost medieval: front-line roads are roofed in nylon nets and barbed wire to tangle the drones, and Ukraine plans thousands of miles of it. This is a war of algorithms, answered with chicken wire.

But tethering and netting are patches on the same fragile thing: the human on the end of the link. Remove the human, and the machine improves — autonomous navigation, CSIS researcher Kateryna Bondar found, lifts a drone’s success rate from around 10–20 per cent to around 70–80, by deleting the two weakest components, the radio link and the stressed operator. So the programme’s stated centrepiece — five times as many operators by 2027 — runs against its own logic. A strike-drone team needs three to six specialists; Ukraine trained over 5,000 last year and only 437 reached the top grade; and no hand on any joystick can meet a saturation swarm. The operator is not merely expensive. The operator is the bottleneck — and in a system built for speed, bottlenecks get engineered out.

Strategists have a name for where this ends. Chinese military thinkers anticipate a “battlefield singularity” — the point at which combat outpaces human cognition and decisions pass to machines — and warn Beijing may be readier to cross it. The competitive logic is merciless. As Robert Work, the former US deputy defence secretary, put it in Schmidt’s telling: “In the future, if the first person through the door is a human and not a robot, we are very stupid people.” NATO’s side is candid too: its innovation chief, Jean-Charles Ellermann-Kingombe, calls this “a tech race” and notes that Russia now iterates a design in as little as two to twelve weeks — against the twenty-four months NATO’s own Rapid Adoption Action Plan allows itself — before adding, of autonomy: “The way you can remove human beings from harm’s way, why wouldn’t you?”

THE SCEPTICS’ HOUR

The revolution thesis deserves its cross-examination, and serious scholars supply it. If Schmidt is precise mass’s most confident champion, its most formidable doubters are academics: Antonio Calcara and colleagues argue that drones have not revolutionised war at all, because air defence adapts and the old contest between hiding and finding endures. The numbers give them ammunition — by some assessments 60–80 per cent of small attack drones never reach their targets — and even CNAS, a home of the precise-mass argument, judges the change in Ukraine evolutionary rather than revolutionary. Cheap drones have burned bombers but won no one a strategic victory; the Donbas may not translate to the Pacific; and the $40 billion may prove less new money than it sounds, or may flow to the familiar giants rather than the proven newcomers.

All fair. But note what the sceptics concede. As Andriy Zagorodnyuk, Ukraine’s former defence minister, put it in a paper for the Carnegie Endowment, no one claims a single weapon has made all others redundant; what changes in a revolution is “the balance of capabilities” and the organisation built around it. That is exactly the claim here. The tank did not abolish infantry. It abolished the primacy of those who were slow to buy tanks.

So judge Drone Edge by its own tests: whether the adoption clock is met, whether the first contracts reward the proven or the merely established, and whether the cost of an interception ever falls below the cost of the thing intercepted.

In Ankara, the alliance formalised what Ukraine, Iran, and the drone incursions over Poland had already demonstrated: precision has been democratised, and power now belongs to those who can produce it cheaply, reorganise quickly, and automate first. NATO has acknowledged the first two imperatives; the third it is funding almost in silence. It is investing in a new military edge—but whether that edge ultimately belongs to human operators or to increasingly autonomous systems is the question Ankara left unanswered. The answer, however, is already being assembled, one procurement standard at a time. The last human in the loop will not be removed by decree. It will be priced out.

When Giorgia Meloni stood before the press in Ankara on 8 July 2026, the headline most outlets reached for was the number: Italy arrived at the NATO Summit in Ankara with a commitment to spend 2.8% of GDP on defence and security.

But the more important message was not about how much Europe spends.

“If Europe increases what it spends without asking the political question of who controls the fundamental supply chains of defence'“ she warned, it risks paying to finance its own dependence.

That remark deserves more attention than it has received. It reframes the whole debate now travelling under the label “Europeanisation of NATO”—the idea, increasingly taken for granted in Brussels and Washington alike, that Europe must become the Alliance’s “second pillar.”

Much of that debate is still framed in the language of accounting: first the 2% benchmark, then the new 5% target agreed at The Hague in 2025 (3.5% for core defence and 1.5% for broader security, with a review scheduled for 2029). Meloni's intervention shifts the discussion onto different ground. The real question is not how much Europe spends, but what it buys, from whom, and whether that spending delivers strategic sovereignty—or merely a more expensive form of dependence.

This raises two important questions, which this article does not claim to resolve.

First: will the Europeanisation of NATO actually deliver greater European sovereignty, or will it simply create new dependencies presented as strategic autonomy?

Second: if Europeans are asked to shoulder a larger share of the burden, will they also gain more say in decision-making — or will they pay and produce while Washington continues to set the strategic direction?

The Italian position, followed over the past five years, offers a sharper way of asking both questions. It does not provide the answers.

AN EVOLVING LINE, NOT A SUDDEN TURN

The Italian position did not begin with Meloni. At the Brussels summit in June 2021 Mario Draghi framed security as the precondition for democracy and prosperity, and pressed for a “broad-spectrum” approach that kept Mediterranean instability in view alongside the eastern flank. The 2022 Strategic Concept, drafted in the shadow of the invasion of Ukraine, named Russia the most significant and direct threat, and named Africa and the Sahel for the first time in a final communiqué. That was an Italian fingerprint, even if the south remained plainly secondary.

Under Meloni, Italy’s approach has sharpened rather than fundamentally changed. At the 2023 Vilnius summit, she firmly backed NATO’s focus on the eastern flank while simultaneously pressing for greater attention to the southern flank.

In 2024, the United States put forward an action plan for the south and appointed a Special Representative for the region. This sparked a sharp dispute when the post was given to a Spanish diplomat rather than someone who better reflected Italian priorities — a grievance Rome has not dropped.

By the 2025 Hague summit and the 2026 Ankara summit, Italy’s argument had evolved. It was no longer just a plea for more attention to its own neighbourhood. It had become a broader strategic vision of what European defence should be for.

Three main threads run through this entire period. It is worth examining each of them in turn, because together they explain why the two open questions are so difficult to answer.

THE SOUTHERN FLANK of NATO

The first thread is geographic. A Europeanised NATO cannot be an eastern-flank alliance with a European label. Italy’s Multi-year Defence Programming Document for 2025–2027 fixes the “wider Mediterranean”. Gibraltar to the Red Sea—as an area of priority strategic interest, and treats instability not as an episode but as a permanent condition. The concrete proof is operational: Italy now leads NATO’s Task Force X in the central Mediterranean, the first time the Alliance has handed a single nation the lead of such a formation.

This is not narrow southern self-interest. Europe’s vulnerabilities run through energy supply routes, migration pathways, key maritime chokepoints, and raw material flows coming from Africa. The Mattei Plan is the connective tissue here, casting Italy as a bridge between the Alliance and the African continent. The danger the Italian line keeps pointing at is that European threat perceptions remain fragmented: the Baltics and Poland, already spending well above 3.5%, reasonably fixed on Russia; the southern members watching a different and less legible set of risks. A serious Europeanisation has to integrate those perceptions rather than let them compete. The alternative is that NATO reproduces internally the same fragmentation that has long hobbled European foreign policy.

THE SUPPLY CHAINS

The second thread is industrial — and it is where Meloni’s warning in Ankara hits hardest.

She highlighted NATO’s own December 2024 list of twelve critical raw materials. At least six of them, she noted, are controlled by a single country, in some cases up to 70% of global production. The picture for refining and processing is even starker, with China dominating the processing of several key minerals to an even greater degree. On several measures, Italy is the most exposed of the EU’s large economies to these vulnerabilities.

The logic is uncomfortable for anyone who treats defence as a spending line, especially those who consider it the main cause for limited expenditure in health care or tax reduction. If European states spend more but remain dependent on external suppliers for semiconductors, sensors, propulsion systems, drones, ammunition and the raw materials underneath all of them, then Europe becomes more expensive to defend without becoming more autonomous.

Meloni’s blunt illustration — a tank worth millions destroyed by a drone costing just twenty thousand euros — makes a deeper point: defence policy and industrial policy can no longer be treated separately.

Rome has begun to answer institutionally, with a coordinating push on defence-critical materials inside the Defence Ministry and a Rome–Berlin initiative in Brussels for a minerals partnership spanning the G7, Africa, India and South Korea. Whether these amount to sovereignty or merely a rearrangement of dependence is, precisely, the first open question.

SECURITY IS NOT THE SAME AS WEAPONS

The third thread is conceptual, and it is the one I find most clarifying. Security, in the Italian reading, is not the same thing as weapons systems.

It is the entire defence system: critical infrastructure, energy supplies, cyber security, data, and the resilience of society to withstand shocks.

Crosetto has put it as plainly as a defence minister can: security is no longer limited to the traditional military domain. It now extends to banking IT systems, hospitals, and the continuity of democratic institutions.

In his framing, defence means protecting critical infrastructure, the country’s vital networks, and its industrial and technological base. Hybrid warfare, he argues, is ongoing and targets decision-making centres and essential services rather than conventional front lines.

A recent case makes the distinction concrete. In early July 2026, Italian investigators arrested two retired intelligence officials accused of selling classified material, reportedly including information about defence-industrial production, to a Russian intelligence officer operating under diplomatic cover. The incident echoes a 2021 case: a Navy frigate captain was caught passing NATO documents to a Russian official and was later sentenced, across separate military and ordinary trials, to 29 years and to 20 years, both now definitive. The current details are, for now, contested and the matter is before the courts; the accused are presumed innocent, and the sharper specifics come from leaked investigative reporting rather than any verdict. But the shape of the case is the point. The asset compromised was not hardware. It was the integrity of the information and counter-intelligence system itself—exactly the terrain on which a “360-degree” definition of security insists the real contest now takes place. You can hit 5% of GDP and still lose that contest for four thousand euros an item.

RESPONSIBILITY WITHOUT POWER?

This brings us back to the second open question.

Italy’s vision of autonomy is carefully limited: it is not about independence from NATO or the United States, but about Europe taking on greater responsibility *within* the Alliance. Crosetto has described the European Defence Community as “no longer postponable,” yet always as a way to strengthen NATO, never to replace it.
The trouble is that the Alliance is drifting from burden-sharing toward what analysts increasingly call burden-shifting, with Europe taking primary responsibility for its own conventional defence while Washington keeps the nuclear umbrella and pivots elsewhere. The unspoken risk is a further step, burden-shedding, though Italian officials treat that as a contingency to plan for rather than an accomplished fact.

Here the honest analysis is uncomfortable. As ISPI has argued, the enablers that make deterrence credible—command and control, strategic lift, intelligence and surveillance, cyber—still sit disproportionately on the American side. European autonomy is not measured by distance from Washington but by the capacity to reduce operational dependence on it. If Europe builds the forces, stocks and platforms that execute plans still written and enabled by the United States, then it will have acquired responsibility without acquiring authority. And responsibility without authority is not sovereignty. It is a more elaborate dependence.

That is the real test — and it is why neither question can be settled here.

A Europeanised NATO will succeed only if it transforms higher spending into real military capability, capability into genuine sovereignty, and sovereignty into meaningful political responsibility — carrying the actual decision-making weight that the term implies.

The 2029 review will be the moment to measure the numbers against the ambition. Until then, these two questions remain open for discussion in the comments.

Sources:

Meloni at Ankara (8 July 2026): The full press conference and the quote on controlling supply chains.
governo.it/it/articolo/vertice-nato-2026-la-conferenza-stampa-d…

The Hague Investment Plan (2025): NATO’s 5% target agreement (3.5% core + 1.5% broader security).
en.wikipedia.org/wiki/Agreement_on_5%25_NATO_defence_spending_b…

Draghi at Brussels (June 2021): “Security as precondition for democracy and prosperity.”
governo.it/it/articolo/intervento-del-presidente-draghi-al-vert…

NATO Strategic Concept 2022: First mention of Africa and the Sahel in a final communiqué.
nato.int/strategic-concept

Meloni at Vilnius (July 2023): “Greater attention to the southern flank.”
governo.it/it/articolo/vertice-nato-vilnius-conferenza-stampa-d…

Italy’s DPP 2025–2027: Multi-year Defence Programming Document, fianco sud, “permanent condition.”
analisidifesa.it/2025/12/la-programmazione-della-difesa-italian…

Mattei Plan for Africa: Italy as bridge between NATO and the African continent.
governo.it/it/piano-mattei

NATO critical raw materials (2024): December list showing single-country dominance.
Referenced in Meloni’s Ankara remarks (see first link above).

China’s dominance in critical-minerals processing: IEA commentary on supply concentration.
iea.org/commentaries/with-new-export-controls-on-critical-miner…

Italy’s exposure to critical minerals: MilanoFinanza analysis of EU vulnerability.
milanofinanza.it/news/terre-rare-perche-l-europa-si-e-alleata-c…

Crosetto at Trento (May 2026): Security beyond military dimension; banking, hospitals, democratic continuity.
jlcnews.it/post/crosetto-la-sicurezza-non-%C3%A8-un-costo-ma-un…

Crosetto’s Lectio Magistralis: “Defence is protection”; hybrid war strikes decision-making centres.
difesa.it/primopiano/valorizzazione-del-patrimonio-e-della-cult…

2021 espionage case (Walter Biot): Navy Commander, NATO documents, definitive sentences (29 + 20 years).
ilpost.it/2026/05/06/cassazione-conferma-condanna-walter-biot-s…

Crosetto on the European Defence Community: “No longer postponable.”
adn24.it/blog/2026/07/02/crosetto-non-e-piu-rinviabile-la-creaz…

ISPI on European strategic autonomy: The enablers (C2, lift, ISR, cyber) remain American-dominated.
ispionline.it/it/pubblicazione/lautonomia-strategica-europea-si…

The NATO Summit scheduled for July 2026 in Ankara could prove to be one of the most important meetings of the Alliance since the end of the Cold War. Faced with the ongoing Russian-Ukrainian war and increasingly clear signs of shifting strategic priorities, European NATO members confront the need to assume greater responsibility for the continent’s security.

NATO at a strategic turning point

One of the most important concepts currently emerging in the debate on the future of Euro-Atlantic security is the “Europeanisation of NATO”. This concept, sometimes referred to as “NATO 3.0”, envisages adapting the Alliance to new geopolitical realities by increasing the role of European states in deterrence, defence, and strategic leadership, while maintaining the transatlantic nature of the organisation. This shift is based on a process of burden-sharing, in which European allies and Canada are expected to assume a leading role in the conventional deterrence and defence of the continent, while the United States would primarily provide the nuclear umbrella and key enabling capabilities. Ankara could therefore become the venue for the official political confirmation of this process.

The “America First” doctrine and the transformation of transatlantic ties

NATO’s Europeanisation is a response to growing tensions in transatlantic relations. These tensions stem largely from both the actions of the United States itself and profound divergences in threat perceptions between Washington and European capitals. While Europe views Russia as an existential threat, the US National Defense Strategy now describes it only as a “persistent but manageable threat”, no longer perceiving it as a major military challenge. The United States is increasingly focusing on the defence of the Western Hemisphere, in line with the “America First” doctrine, as well as on issues such as security in the Indo-Pacific, economic and technological competition with China, and efforts to contain Iran’s nuclear programme.

However, it is worth noting that significant differences in threat perceptions also exist within Europe, which translate into inconsistencies in its approach to cooperation with Washington.

- Eastern and Northern European countries (including the Baltic states) prioritise deterrence against Russia and advocate for detailed operational planning in the event of a direct attack.

- Southern European states are increasingly focused on instability in the Mediterranean and the Middle East, as well as the challenge of irregular migration.

- For countries such as Türkiye, domestic and cross-border terrorism remains a key security concern.

Managing these fundamental differences – particularly the fact that the United States and Europe no longer share a common understanding of the Alliance’s primary adversary – is widely regarded as the most significant source of current tensions within NATO.

The shift in European policy is also closely linked to growing concerns about Washington’s political unpredictability, including Donald Trump’s threats to withdraw from the Alliance, the conditioning of security commitments on political objectives (for example, support for US policy in the Middle East), and unilateral decisions regarding the withdrawal of American troops from Europe, including Germany. Such uncertainty undermines the credibility of deterrence, making reliance on US security guarantees increasingly problematic from a European perspective.

This strategic shift, combined with the US administration’s transactional approach to the Alliance, is contributing to a gradual reduction of the American military presence in Europe.

Among other measures, the Pentagon has announced the withdrawal of 5,000 troops from Germany, the suspension of plans to deploy 4,000 troops to Poland, the cancellation of conventional cruise missile deployments in Germany, and a permanent reduction in the number of Brigade Combat Teams (BCTs) assigned to Europe from four to three. By 2027, it is expected that the United States may be able to provide only half – or potentially even less – of the Alliance’s overall combat capabilities.

As a result, public sentiment across Europe has begun to shift. Declining trust in the United States is increasingly reflected in public opinion.

Polls indicate that more than 70% of Europeans believe that, after decades of close cooperation with Washington, Europe should prioritise greater strategic autonomy or, at the very least, diversify its security partnerships as much as possible. In this context, Europeanisation is not an attempt to exclude the United States from the Alliance, but rather a pragmatic contingency plan and a means of safeguarding European security in the event that US support during a crisis proves limited, delayed, or constrained by political considerations.

Consequences of reducing the US military presence

The changes taking place in US strategy have not only political but also military dimensions.

In recent years, reports have emerged of a reduction in the US military presence in Europe, a decrease in the number of brigade combat teams, and a revision of plans for the deployment of new missile systems.

According to numerous analyses, after 2027, the United States may only be responsible for approximately half of NATO’s combat capabilities. This would constitute the largest change in the Alliance’s structure since its inception.

For European countries, this means the need to independently supplement their capabilities and create their own political and military command structures, which Washington has provided for decades. This process includes several key elements:

- increasing defence spending to a level exceeding 5% of GDP;

- Europeans assuming greater responsibility for NATO’s operational command;

- developing a European defence industry;

- increasing Europe’s share in financing military support for Ukraine;

- building its own conventional deterrent capabilities.

The E5 format as the basis for NATO’s “European pillar”?

The response to the erosion of American guarantees is the construction of a European pillar within NATO, the architecture of which is based on the emerging balance of power on the continent. The E5 format, bringing together the continent’s largest military powers – France, Germany, Great Britain, Italy, and Poland – plays a key role in shaping the new security structure.

The emergence of this initiative, however, is not a random development but a deliberate one. A growing number of European capitals are realising that the European Union’s decision-making mechanisms are too slow and inefficient, due to the need for collective approval, to respond effectively to the rapidly deteriorating security environment.

In light of these trends, speculation regarding the potential evolution of the E5 format is increasingly emerging. Some experts believe that this format could ultimately evolve into an informal European Security Council, bringing together the countries with the greatest military potential and the ability to make strategic decisions quickly. This structure would help overcome some of the procedural constraints of the European Union while creating a space for cooperation with non-EU countries, particularly the United Kingdom, Norway, Canada, and Türkiye. The creation of such a forum could significantly influence the future shape of NATO’s European pillar and strengthen Europe’s ability to respond independently to security threats.

This process has already reached an operational level, with European officers taking over command of NATO’s three regional Joint Force Commands from American officers. Preparatory discussions within the Alliance are increasingly focused on realising the vision of a “European-led and European-owned NATO”. The most significant challenge, however, remains the creation of effective political and military command structures capable of managing potential conflict escalation without having to rely on Washington’s decisions.

Operational gaps: ELSA, deep strike, and the subscription model

The war in Ukraine has demonstrated that the ability to strike targets hundreds of kilometres behind the front lines is crucial to effective warfare. However, the current capabilities of many European countries remain limited. The Ankara Summit will force them to confront the capability gaps that have emerged in European arsenals as a result of their long-term reliance on American technology.

One of the greatest weaknesses is long-range precision strike capability. Currently, European conventional systems are limited to a range of approximately 500 kilometres, leaving Russian operational and logistics bases beyond their reach. To address this challenge, the ELSA initiative (European Long-Range Strike Approach) was established, a programme focused on building European capabilities to conduct long-range precision strikes (so-called deep strikes). ELSA is intended to enable the development of a European “strike chain”, encompassing sensors, reconnaissance systems, intelligence analysis systems, and weapons capable of striking deep into enemy rear areas.

While this is a promising step toward closing this gap, the project currently suffers from an excessive focus on platforms (i.e., acquiring missiles) while neglecting system integration. Simply acquiring missiles is not the same as building a comprehensive strike system. For ELSA to truly strengthen Europe operationally, it must integrate modern sensors, intelligence fusion systems (ISR), and real-time decision-making mechanisms. Without this, cutting-edge missiles will be merely “expensive signalling devices” rather than tools capable of shaping the battlefield. To build these and other key capabilities – including naval systems, unmanned aerial vehicles, and air defence systems – European countries must drastically increase funding. The current threshold of 2% of GDP for defence has become insufficient and therefore obsolete, which is why the new benchmark is 5% of GDP.

Equally important is the reform of the military procurement process itself. The introduction of a so-called “subscription” model, based on multi-year framework agreements rather than one-off tenders, is intended to provide arms manufacturers with predictable demand and enable continuous production. Such a solution would keep the defence industry constantly prepared and enable the faster deployment of new technologies on the battlefield. Ukraine’s experience has shown that the advantage often goes not to those with the most expensive weapons systems, but to those able to produce them faster and in larger quantities.

Is a European pillar of NATO emerging?

The most important question today is not whether the Europeanisation of NATO will occur, but how quickly it will proceed.

Many indications suggest that Europe is gradually building its own security pillar within the Alliance. The development of the E5 format, the ELSA initiative, increased defence spending, and the growing role of countries like Poland and Türkiye are all part of the same process.

The greatest obstacle, however, remains the lack of unified European political leadership and mechanisms for rapid strategic decision-making.

The NATO Summit in Ankara could become the moment when European states officially accept the need to take greater responsibility for their own security. If this happens, Ankara 2026 will be remembered not only as another Alliance summit, but as the beginning of a new phase in NATO’s history—a phase in which Europe ceases to be merely a recipient of security and becomes its main co-creator.

When Iranian drones hit Amazon data centers in the UAE and Bahrain back on March 1, 2026, the disruption hit hard and fast.

Payment services like Alaan and Hubpay went down, banks such as Emirates NBD and Abu Dhabi Commercial Bank saw their mobile apps and phone banking fail, and even Careem — the go-to ride and delivery app across the region — struggled to keep up. The attack is believed to be the first time data centers have been deliberately targeted for air strikes in a conflict.

While these temporary service interruptions were significant, the attack’s deeper implications extended far beyond short-term downtime. It exposed the physical vulnerability of the very infrastructure underpinning the Gulf’s ambitious push for economic diversification into data centers, cloud computing, and artificial intelligence (AI). In an era where digital assets are touted as the foundation of post-oil prosperity, the strikes underscored how 21st-century economic power remains tethered to fragile, physical target.

In the 21st century, the world’s most valuable infrastructure turns out to be surprisingly fragile when missiles start flying.

The Massive AI Push in the Gulf

Lets have a look at what’s been happening over the past couple of years: the Gulf has become a major player in the global AI boom. Microsoft is pouring about $15 billion into the UAE by 2029. Amazon is building out an AI hub in Riyadh. Nvidia is teaming up with Saudi Arabia’s Humain to deliver hundreds of thousands of GPUs. And then there’s Stargate UAE — a huge collaboration between OpenAI, Oracle, G42, and others that could become one of the largest AI campuses anywhere outside the United States.

Additional momentum comes from Blackstone’s roughly $3 billion partnership with Humain for Saudi data centers and broader collaborations involving MGX, the UAE’s AI investment vehicle.

These projects leverage the Gulf’s natural strengths: its strategic location at the crossroads of Europe, Asia, and Africa positions it as an ideal hub for low-latency data traffic serving billions, while abundant, low-cost energy—far more reliable than strained U.S. grids—provides a competitive edge for power-hungry AI workloads.

GCC capitals, led by Riyadh and Abu Dhabi, have channeled hydrocarbon revenues into hyperscale data centers and AI clusters. Sovereign wealth funds totaling nearly $6 trillion back this shift. According to PwC, AI could contribute approximately 12.4% to Saudi Arabia’s GDP and nearly 14% to the UAE’s by 2030, driving substantial economic gains. Saudi Arabia’s Public Investment Fund has established vehicles targeting tens of billions for AI, while the UAE created a dedicated Ministry of Artificial Intelligence. These efforts treat computing infrastructure as a multi-decade national project for digital sovereignty and diversification.

Emerging Vulnerabilities: Physical Threats and Tech Dependencies

This capital-intensive drive toward diversification, however, confronts two critical vulnerabilities: escalating geopolitical threats that now target physical infrastructure and a persistent reliance on Western (particularly U.S.) technology supply chains.

For years, risks to digital systems centered on cyberattacks—ransomware, intrusions, and exploits. The March 2026 strikes marked a shift, demonstrating that precision drones and missiles can directly impair commercial data centers, much like earlier attacks on oil facilities (e.g., the 2019 Aramco strikes). Iran’s influence over the Strait of Hormuz and its capacity for further strikes amplify these concerns, turning data centers into strategic targets in regional conflicts.

The supply chain challenge is equally profound. While Gulf states supply capital, land, and energy, core technologies—advanced GPUs, proprietary software, and cloud architectures—originate primarily from the U.S.

MGX’s partnerships with BlackRock, Microsoft, and others (including the AI Infrastructure Partnership aiming for tens of billions in data center investments) exemplify this dynamic.

The United States retains a clear advantage in cutting-edge semiconductors, but stringent export controls often complicate and delay access to these technologies. U.S. export controls on high-end hardware, managed by the Department of Commerce, give Washington leverage. China, by contrast, offers less advanced chips but imposes far fewer export restrictions and has emerged as a leader in open-source AI models. This combination makes Chinese technology an attractive option for Gulf governments seeking to rapidly develop domestic AI infrastructure and capabilities.

From Oil Dependency to New Interdependencies

The Gulf's strategy aims to reduce dependence on hydrocarbon revenues by redirecting capital toward strategic domestic investments and productive assets.

Yet, the pivot to AI has traded one form of dependency for another. Instead of full autonomy, the region’s digital future hinges on regional security, stable power grids, and foreign regulatory approvals for chips and updates.

Additional examples illustrate this paradox. The Stargate UAE project, while transformative, relies on U.S. partners for operations and faces direct threats from Iran, which has signaled targets among Western tech firms and Gulf countries. As “compute becomes the new oil,” Gulf data centers inherit the strategic risks once reserved for energy infrastructure, creating asymmetric interdependencies where the U.S. gains from Gulf energy and capital, but Gulf ambitions depend on American technology and security umbrellas.

Looking Ahead

The Gulf’s post-petroleum sovereignty drive is yielding impressive infrastructure and growth projections.

Gulf states will continue to be central to the landscape of global computing power with or without the United States.

On the other hand, true diversification requires not only massive investment but also enhanced physical defenses, supply chain resilience (perhaps through greater localization or diversified partnerships), and diplomatic efforts to stabilize the region. The future of Gulf AI leadership will depend on addressing these new dependencies head-on—transforming ambitious visions into robust, secure realities in a geopolitically volatile environment.

As the academic year comes to a close, Oxford is slowly beginning to change its rhythm. Students are gradually leaving the city, college courtyards are becoming quieter, and the familiar pace of term time is giving way to the lighter atmosphere of summer. At the same time, tourists are filling the streets, cameras in hand, bringing a different kind of energy to Oxford’s historic lanes, colleges, and libraries.

It is in this changing summer atmosphere that we, at the Oxford Geopolitics Initiative, are looking back on an active and productive Trinity Term.

Oxford’s academic year is traditionally divided into three terms: Michaelmas, Hilary, and Trinity. At the beginning of Trinity Term, we set ourselves an ambitious agenda: to organise high-quality events, bring together diverse voices, and publish timely analysis on some of the most pressing geopolitical questions of the day.

The Oxford Geopolitics Initiative’s research objectives for 2026 focused on three key themes: the future of AI, geopolitics, and the global divide; the rise of civilian technology as strategic power in a new era of dual-use innovation and global competition; and the shifting balance of power among China, middle powers, and the West.

A Term of Events, Discussions, and New Ideas

During Trinity Term, the Oxford Geopolitics Initiative hosted and contributed to a series of events across the University of Oxford. These included discussions on artificial intelligence, European security, social entrepreneurship, food security, and the future of industrialisation.

Some of our key events included:

The Geopolitics of AI: Will Algorithmic Efficiency Shatter the US-China Duopoly?

Dr Robert Hercock is a leading UK researcher in Cyber Security and AI. Held at Pembroke College, this event explored how algorithmic efficiency may reshape the global AI race and challenge existing assumptions about technological competition between the United States and China.

Can Türkiye Be a Partner for European Security?
Hosted at the Blavatnik School of Government, this discussion examined Türkiye’s evolving role in European security architecture and its strategic importance in a changing geopolitical environment.

How to Industrialize in the 21st Century?
In collaboration with the University of Oxford’s TIDE Centre — the Technology and Industrialisation for Development Centre — we discussed what industrialisation looks like in the twenty-first century and how countries can pursue development in an era of technological disruption, shifting supply chains, and geopolitical uncertainty.

AI and the Future of Global Food Security
This event addressed the relationship between emerging technologies and global food systems, with a particular focus on how AI may affect food security in the years ahead.

Publishing Timely Geopolitical Analysis

Alongside our events, we continued to develop our Substack as a platform for thoughtful, timely, and accessible geopolitical analysis.

Over the course of the term, we published more than ten pieces addressing major international developments and long-term strategic questions. These included:

Is the US at War with Iran — or with Geography? by By Onur Kovaci, Kings College

Connecting the Two Ends of Eurasia: The Trans-Caspian Corridor and Implications for the South Caucasus, Mongolia, and Japan By Shu Uchida, PhD, University of Tokyo

Can Britain Afford Rearmament? Between Strategic Ambitions and Fiscal Constraints By Natalia Adrianna Potera, Defence24.com

The Global AI Split: China, the U.S., and Everyone Else, By Seyithan Ahmet Ates, Pembroke College, Oxford University

What the Iran-US Negotiations Need?, by Mohammed Alharthi, Blavatnik School of Governance, Oxford University

Our 2026 cohort has also been highly active. Together, they have published more than 20 analyses on Substack, contributing to our growing platform with fresh perspectives, regional expertise, and critical insight. Their work will continue throughout the summer.

Members of our fellowship community have also represented the Oxford Geopolitics Initiative beyond Oxford, participating in external forums, policy discussions, and international events.

Looking Ahead to the Summer

Although Oxford is entering the summer vacation period, our work will continue uninterrupted.

In the months ahead, we will continue to organise online events, publish analytical work, and engage with the most pressing questions shaping international affairs. From great-power competition and AI governance to regional security, development, and the global political economy, we will remain attentive to the issues that matter most.

We are grateful to our speakers, contributors, fellows, readers, and audience members for being part of this journey.

We look forward to seeing many of you at our in-person events in Oxford next term, as well as at our online meetings over the summer.

Until then, we wish you a pleasant and restful summer.

Greetings from Oxford,
Oxford Geopolitics Initiative

Russian aggression in Ukraine, the deteriorating security situation in Europe, growing instability in the Middle East, and the gradual reduction of American involvement in European security are forcing London to redefine its defence policy. At the same time, the country is grappling with a difficult budgetary situation, high debt, rising social spending, and pressure to maintain public service funding. This raises the fundamental question of whether the UK can finance the rebuilding of its military capabilities without incurring high political and economic costs.

This debate centres on the long-awaited Defence Investment Plan (DIP), which will define how British defence will be financed over the coming decade. The document, whose publication was repeatedly delayed, has become symbolic of a wider problem in British security policy: the mismatch between the country’s declared ambitions and its actual capabilities.

The Crisis of British Defence Policy

For many years, British defence policy operated under conditions of structural underfunding. Successive strategic reviews assumed that the UK could maintain its global role while simultaneously cutting spending and reducing the size of its armed forces. In practice, this led to a gradual erosion of the country’s military potential.

The scale of these changes is significant. In 1990, the British Army numbered over 153,000 professional soldiers. Today, there are fewer than 74,000. The number of reservists has fallen from around 76,000 to fewer than 26,000. The Royal Navy has reduced its number of major combat ships from 48 to just 13, while the Royal Air Force currently operates significantly fewer combat aircraft than it did three decades ago.

Although modern weapons systems are more technologically advanced, the reduction in force size has resulted in limited capabilities for sustained operations and simultaneous responses to multiple crises. As a result, an increasing number of analysts argue that the UK maintains only a „skeleton” of its full military potential. This means that it possesses almost all types of forces and systems, but in numbers insufficient to conduct large-scale military operations.

Great Power Ambition Versus Reality

One of the main issues that can be identified is the persistent gap between London’s strategic ambitions and the country’s actual resources. British decision-makers continue to view the UK as a key Western military power, capable of playing a leading role in NATO and participating in global operations.

In recent years, London has considered, among other things:

- leading security missions following the withdrawal of US troops from Afghanistan

- participating in a potential peacekeeping force in Ukraine

- securing naval operations in the Strait of Hormuz

- fielding significant land forces for NATO

However, many analysts believe that the potential of the British armed forces no longer matches these ambitions and that any foreign mission now requires deploying military resources to their limits, or even beyond. A good example of this occurred during the Middle East crisis in 2026, when the UK needed several weeks to deploy a single destroyer to the Persian Gulf. This event became a symbol of the limited power-projection capabilities that, just two decades ago, formed the foundation of British global strategy. This apparent gap between Britain’s ambitions and the resources it actually possesses indicates a high degree of self-deception regarding its actual military capabilities.

Ultimately, the DIP will have to choose between focusing on the ability to defend the UK and its immediate surroundings or restoring its global power-projection capabilities—a choice that appears extremely risky, both politically and economically.

The End of the Era of Total Dependence on the United States

For decades, British security strategy was based on the assumption that the United States would remain the primary guarantor of European security and that London would play the role of Washington’s most important partner on the continent.

Today, this assumption is increasingly being questioned. A report by the British House of Lords emphasized that the high level of military dependence on the United States is no longer a sustainable foundation for British strategy. The need to build more autonomous defence capabilities and deepen cooperation with European partners is being discussed increasingly frequently. The problem, however, is that greater strategic autonomy requires significant financial investment and the reconstruction of industrial capacity, which has been neglected and marginalised for years.

The Defence Investment Plan and the Financing Problem

The key document aimed at addressing these problems is the Defence Investment Plan. According to press reports, this plan is expected to envisage investments of approximately £15 billion, financed primarily through spending cuts in other sectors of public administration.

At the same time, the Ministry of Defence indicates that the funding gap in current defence plans is approximately £28 billion. Even after partial deficit reduction, the shortfall remains around £18 billion, and the Treasury is only prepared to cover a fraction of this amount.

The situation was further exacerbated by the political crisis triggered by the resignations of Defence Minister John Healey and Armed Forces Minister Al Carns. Both politicians openly accused the government of underfunding defence and lacking the willingness to make difficult budgetary decisions. Healey even stated that the proposed plan „does not meet the country’s defence needs at a time of growing threats.”

Who Will Pay for the Military?

The biggest challenge remains finding the necessary funding. Prime Minister Keir Starmer has committed to gradually increasing defence spending to 3% of GDP and, in line with emerging NATO targets, potentially to 3.5% of GDP for defence and 5% of GDP for broader security-related spending. Currently, the UK spends around 2.3% of GDP on defence.

This would require finding tens of billions of pounds in additional funding each year. According to estimates, achieving the 3.5% target alone could require an additional £36 billion annually.

Resolving the issue of defence funding would be particularly desirable before the July NATO summit in Ankara, where US President Donald Trump will undoubtedly reiterate his call for allies to increase defence spending and reduce their reliance on American security guarantees.

In practice, the British government faces three options:

1. Cuts in public spending – including development aid, climate policy, transport, and certain social programmes.

2. Increased public debt – a solution the Treasury remains reluctant to accept.

3. Tax increases – a politically costly option for the Labour government.

It is becoming increasingly clear that higher military spending will not be possible without significant compromises in other areas of public policy. While many people understand the need to strengthen national defence, this may come at the expense of public services, particularly healthcare, potentially affecting the overall quality of life and well-being of the average Briton.

The UK, NATO, and European Defence Autonomy

The debate over British defence funding is not limited solely to the national budget, which alone would be unable to cover the growing costs associated with the current need for rearmament. Therefore, the UK’s role in the evolving European security architecture is becoming increasingly important. A particularly significant example is the dispute over the UK’s participation in the SAFE (Security Action for Europe) programme, a €150 billion fund aimed at supporting joint defence procurement among European states. Cooperation could also be extended to selected partner countries, which the UK could theoretically join. (As of June 2026, Canada is the only third country to have concluded a full agreement with the EU under Article 17 of the SAFE Regulation.)

However, despite political declarations regarding the need to deepen defence cooperation between the EU and the UK, no formal agreement has yet been reached on British participation in SAFE. The main sticking points remain the size of London’s financial contribution and the extent of access for British companies to contracts financed under the programme. A lack of agreement on participation fees led to the breakdown of negotiations at the end of 2025.

Currently, British companies can participate in SAFE-funded projects only on a limited basis, which, from London’s perspective, is insufficient. The UK is seeking a greater role, arguing that it possesses one of the largest and most advanced defence industries in Europe.

Minister for the Armed Forces Al Carns described the lack of agreement between London and Brussels as a „suicidal” step from the perspective of European security. He argued that the UK remains a key pillar of European security and that limiting the British defence industry’s access to new funding mechanisms weakens the defence capabilities of the entire continent.

Countering Russia’s „Shadow Fleet”

The growing hybrid threat posed by Russia provides another argument in favour of increasing defence spending in Europe, including in the U. On 14 June 2026, British special forces conducted the first-ever operation to seize a Russian tanker belonging to the so-called „Shadow Fleet” in the English Channel. Royal Marines commandos boarded the tankerSmyrtos, supported by helicopters, Royal Navy ships, and a maritime patrol aircraft, after which the vessel was placed under the supervision of the British authorities.

The operation forms part of a broader effort by European countries to counter Russian attempts to circumvent sanctions and demonstrates NATO’s growing determination to address hybrid threats. It also highlights that, despite financial challenges, the UK still possesses the capabilities required to conduct complex military operations of strategic importance.

The Problem Is Not Just About Money

Some experts argue that increasing the budget alone will not solve all of Britain’s defence problems. The British military procurement system has for years been plagued by delays, cost overruns, and shortcomings in the implementation of modernisation programmes.

The National Audit Office has repeatedly highlighted problems in the delivery of major defence projects. A significant number of key programmes have been assessed as being at risk of failure, while the acquisition process for new equipment often takes many years.

Furthermore, questions have been raised about the direction of modernisation. Following his resignation, Al Carns argued that some of the planned spending remains focused on systems designed according to the logic of previous conflicts, prioritising highly advanced—and therefore expensive—equipment that is difficult to replace quickly. By contrast, the experience of the war in Ukraine has demonstrated the importance of rapidly developing drones, autonomous systems, and modern battlefield technologies that are effective without necessarily being the most expensive.

Conclusions

The debate surrounding the UK’s Defence Investment Plan (DIP) demonstrates that the UK now faces a strategic choice. It is no longer just about the size of the defence budget, but rather about what role the country wants to play in the international security system and whether it will be able to bear the costs associated with these ambitions.

The growing threat from Russia, uncertainty about the future involvement of the United States in Europe, and an increasingly unstable international environment are forcing European countries to increase military spending. At the same time, fiscal constraints, public pressure, and years of neglect in the modernisation of the UK’s armed forces make the realisation of these ambitions exceptionally difficult.

Ultimately, the most important challenge for the British authorities will not be increasing defence spending itself, but aligning strategic ambitions with the country’s actual capabilities. If the DIP fails to resolve this fundamental contradiction, it may become yet another document promising more than the UK can deliver.

Abstract

The geopolitical crisis of 2022 accelerated a structural rerouting of Eurasian trade, elevating the Trans-Caspian International Transport Route (TITR), or Middle Corridor, as a strategic alternative connecting East Asia to Europe via Kazakhstan, the Caspian Sea, Azerbaijan, Georgia, and Türkiye. This paper examines the TITR from two angles: Japan’s growing engagement with continental Eurasian connectivity — reflecting an imperative to diversify supply chains beyond the maritime Indo-Pacific — and the South Caucasus as an indispensable interface enabling the corridor to function as a transcontinental route. The paper further considers implications for Mongolia as a landlocked state seeking strategic value within Eurasian networks.

Keywords: TITR, Middle Corridor, Trans-Caspian, Japan, South Caucasus, supply chain resilience, Eurasian connectivity, Azerbaijan, Georgia

Introduction

The architecture of Eurasian connectivity is being reconfigured. For decades, overland trade between East Asia and Europe relied predominantly on the Trans-Siberian Railway running through Russian territory. Maritime routes through the Suez Canal served as the primary arteries of global commerce. Both options, however, have been subject to increasing disruption: The geopolitical crises of recent years have rendered the Northern Corridor politically and commercially untenable for many shippers, while maritime chokepoints remain vulnerable to conflicts, insurance surcharges, and geopolitical leverage.

Into this environment, the Trans-Caspian International Transport Route (TITR) — also referred to as the Middle Corridor — has attracted renewed and serious attention. The route links China and East Asia to Europe through Kazakhstan, a Caspian Sea crossing, Azerbaijan, Georgia, and onward to Türkiye or the Black Sea. It is a multimodal corridor, combining rail, maritime, and road segments, and its operation requires coordination among customs authorities, port operators, railway companies, and logistics providers across multiple sovereign states.

This paper proceeds in three analytical steps. First, it examines why Japan — an island state long defined by its maritime orientation — has developed a strategic interest in a continental overland corridor. Second, it analyses the South Caucasus not as a passive transit zone, but as an active and indispensable interface within the TITR architecture. Third, it considers the implications of this analysis for Mongolia, a landlocked state that faces analogous structural challenges and for which the South Caucasus experience may offer a useful, if partial, reference point.

Japan and the TITR: From Maritime Power to Continental Engagement

Japan’s engagement with the TITR represents a significant dimension of its evolving foreign economic policy. Japan is not a continental power in the traditional sense. Its economic model has historically depended on stable sea lanes, imported energy, and maritime supply chains connecting it to global markets. The Free and Open Indo-Pacific (FOIP) vision, which has guided Japanese foreign policy since 2016, is fundamentally maritime in its framing.

Yet, Japan has increasingly recognised that a purely maritime orientation leaves it exposed. The COVID-19 pandemic demonstrated the fragility of just-in-time supply chains. The geopolitical crisis in 2022 disrupted not only the Northern Corridor but also air routes over Russian territory that Japanese carriers had long relied upon for Europe connections. These disruptions have created a strategic case for diversification that goes beyond any single sector or region.

Japan’s engagement with the TITR was formalised at the first Summit of the “Central Asia plus Japan” Dialogue (CA+JAD), held in Tokyo in December 2025. The Summit’s Tokyo Declaration explicitly identified Central Asia’s role as a “junction between Europe and Asia” and committed Japan to supporting the further development of the TITR. This political framing is significant: it signals that Japan views the Middle Corridor not merely as a logistics project but as a structural element of a more resilient and diversified Eurasian connectivity architecture.

The data support this strategic turn. Japanese and South Korean cargo has increasingly been routed through the TITR using Container Rail Express (CRE) services connecting East Asia and Europe. In early 2025, Japanese cargo was successfully transported through the corridor on a trial basis, demonstrating its operational feasibility. According to the TITR Association, cargo volumes transported through Kazakh and Azerbaijani ports reached approximately 3.3 million tons in 2024, a 20 percent increase from the previous year, while container traffic reached 56,500 TEU — a 176 percent increase year-on-year. The World Bank’s 2030 Middle Corridor Action Plan projects that full implementation of planned investments could halve transit times and triple freight volumes by the end of the decade.

For Japan, the TITR offers three distinct strategic benefits. First, diversification: it provides an additional continental option connecting Japan and East Asia with Europe and Central Asia, reducing dependence on any single route. Second, resilience: the existence of an alternative route creates redundancy in the system, protecting against disruption to maritime or northern land routes. Third, continental engagement: the TITR enables Japan to extend its strategic presence into Eurasia through infrastructure partnership, logistics cooperation, and investment — without requiring a confrontational posture. It should be noted that the TITR remains more costly than maritime alternatives, with freight rates running significantly higher than Suez Canal routes; yet for time-sensitive cargo, supply chain diversification, and geopolitically motivated rerouting, this cost premium is increasingly accepted as a strategic investment rather than a commercial inefficiency.

This last point deserves emphasis. Japan’s approach to TITR engagement is deliberately calibrated. The language used in official documents emphasises connectivity, resilience, and stable supply chains rather than geopolitical competition. This framing allows Japan to work cooperatively with Central Asian states, South Caucasus states, Türkiye, and European partners, while avoiding the zero-sum logic that might otherwise constrain its options.

The South Caucasus as Interface: Beyond Transit

If Japan’s engagement illustrates the demand side of the TITR’s growing importance, the South Caucasus illustrates the indispensable structural role that a geographically compact and politically complex region plays in enabling the corridor to function.

After goods cross Central Asia and traverse the Caspian Sea, they reach Azerbaijan, from where the route continues through Georgia toward Türkiye and the Black Sea. This geography is not incidental. The South Caucasus occupies a narrow but structurally critical space between the Caspian and the Black Sea — the only viable overland link between Central Asia and Europe that does not pass through Russia or Iran.

The region hosts the key infrastructure nodes of the corridor: the Baku-Tbilisi-Kars (BTK) railway, which operationalised the rail link between Azerbaijan, Georgia, and Türkiye; the Baku-Alyat port complex on the Caspian Sea, which serves as the primary gateway from Kazakhstan; Georgian railway networks connecting to Black Sea ports at Batumi and Poti; and the BTC oil and BTE gas pipelines, which established an earlier but analogous logic of South Caucasus connectivity. Taken together, these elements constitute not merely a transit zone but what this paper terms a connectivity space — a region whose strategic value lies in its capacity to link adjacent zones that would otherwise be disconnected.

The South Caucasus is also, of course, a region of considerable political complexity. Georgia has pursued European Union (EU) association and is currently navigating a contested political environment. Armenia and Azerbaijan concluded a devastating war over Nagorno-Karabakh, and peace negotiations remain fragile. Russia, Türkiye, Iran, and the EU all maintain competing interests in the region. The result is a layered geopolitical environment in which infrastructure and politics interact in real time.

This complexity, however, does not diminish the region’s importance — it amplifies it. Eurasian connectivity does not depend only on physical infrastructure. It depends on whether difficult regions can remain open, functional, and reliable. The South Caucasus demonstrates, imperfectly but instructively, that a geographically constrained region with contested politics can nevertheless serve as a systemic enabler of wider integration, provided that sufficient institutional coordination, investment, and political stability are maintained.

It is worth noting that the analytical question for the TITR is not which individual state within the South Caucasus matters most. Such framing risks encouraging zero-sum competition and misses the systemic logic of the corridor. The South Caucasus functions as an interface precisely because multiple states — Azerbaijan, Georgia, and potentially, in the future, Armenia — together provide the redundancy, infrastructure, and territorial continuity that the route requires. A peace process between Armenia and Azerbaijan may over time diversify the routing options within the region; this would not reduce the South Caucasus’s importance but confirm it by expanding its connectivity potential.

Transit time reductions further underscore this systemic importance. Compared to maritime routes, transit times have been reduced from 38–53 days to 18–23 days between East Asia and Europe according to the TITR International Association and EY Kazakhstan (2024). This improvement reflects investments across the entire corridor, but the South Caucasus segment has been particularly important given its historical bottlenecks in port capacity and customs processing.

Implications for Mongolia

Mongolia occupies a structural position that is in some respects analogous to, and in others quite different from, that of the South Caucasus. Like the South Caucasus states, Mongolia is a geographically constrained state located between major powers — in Mongolia’s case, Russia and China — that must think carefully about connectivity, transit options, and economic resilience. Unlike the South Caucasus, Mongolia is not currently part of the TITR’s operational network in a direct sense.

The relevance of the TITR for Mongolia is therefore not primarily operational but conceptual and strategic. The experience of the South Caucasus offers Mongolia a reference point for understanding how geographically constrained regions can increase their strategic value within wider Eurasian connectivity networks. The lesson is not to copy the South Caucasus model — the geographic and political contexts are sufficiently different to preclude direct replication. The lesson is rather that transit spaces can acquire significant strategic value when they offer reliable, diversified, and well-managed routes connecting adjacent regions; and that this value depends not only on physical infrastructure but on institutional coordination, customs modernisation, investment climate, and political stability.

Mongolia’s “Steppe Road” policy and its integration with China’s Belt and Road Initiative suggest an awareness of this opportunity. This reflects the growing recognition that Mongolia’s geographic position, while challenging, can be transformed into a strategic asset if the right investments and institutional frameworks are put in place.

Conclusion

The TITR is no longer a political aspiration. It is an increasingly operational and strategically significant Eurasian connectivity space, whose growth is driven by geopolitical disruption, institutional investment, and the structural demand for diversified trade routes. Japan’s growing engagement with the corridor reflects a deliberate strategic evolution, extending its connectivity interests from the maritime Indo-Pacific to the continental Eurasian space. The South Caucasus, meanwhile, functions not merely as a transit region but as an indispensable interface within a wider Eurasian connectivity space. Its significance lies not in the dominance of any single state, but in its capacity to sustain linkages between Central Asia, the Black Sea region, Türkiye, and Europe.

The broader insight this analysis offers is a relational one: the future of Eurasian connectivity will not be determined by the dominance of any single country or route. Rather, it will depend on whether multiple connectivity spaces remain linked, open, and functional. The TITR illustrates how infrastructure, institutions, and geography can together create resilient connections across Eurasia. For Mongolia, and for other geographically constrained states seeking strategic relevance in a reconfiguring Eurasian order, this insight carries both practical and strategic significance.

Note: Dr Uchida’s participation in the UBD was funded by ROLES at RCAST, the University of Tokyo, under a research grant from the Ministry of Foreign Affairs of Japan

About the Author:

Shu Uchida, PhD

Project Researcher, Research Center for Advanced Science and Technology (RCAST), University of Tokyo

‘In the second half of the 20th Century, it was the race to develop nuclear arms that occupied some of the finest minds in the US and the Soviet Union.’

Glenny and Mintz put it well when they described the AI race between China and the US. However, this competition may have wide-ranging consequences for the rest of the world.

A growing body of evidence suggests that China and the United States are consolidating structural advantages across the AI value chain: talent, computing capacity, capital, IPs, frontier models, cloud infrastructure, and industrial ecosystems.

Together, they employ around 70 percent of the world’s top machine-learning researchers and command roughly 90 percent of global computing power. They also attract the overwhelming majority of AI investment, receiving more than twice the combined total of all other countries.

The imminent risk is that in earlier technological revolutions, countries outside the frontier could gradually adopt new technologies and catch up. However, when it comes to AI, it may not allow the same path. Progress in AI depends on advanced chips, data centers, cloud platforms, cheap electricity, and massive capital expenditure. Countries that lack one of these foundations may not simply fall behind. They may become dependent on systems designed, hosted, and controlled elsewhere.

Number of newly funded AI companies by geographic area

The U.S.–China Lead Across the AI Stack

Lets get started with the United States.

The U.S. remains the strongest actor in several decisive areas. It hosts 5,427 data centers, more than ten times any other country, and continues to lead in global private AI investment. U.S. private AI investment is reported to be 23 times larger than China’s, and in generative AI, American investment exceeds the combined total of China and Europe by a wide margin.

China, however, has built a powerful alternative base. It leads in publication volume, citations, and patent grants. The United States still produces more high-impact patents and more top-tier AI models, but China’s scientific and industrial scale is rapidly narrowing the gap.

Number of notable AI models

In 2024, the U.S. produced 40 notable AI models compared with China’s 15. Meanwhile, China’s share of the top 100 most-cited AI papers rose from 33 in 2021 to 41 in 2024. China has also outpaced the U.S. and other countries in AI-related patents by having 25k AI-related patents in 2024, whereas the U.S. has 17k, and the rest of the world around 10k.

China is also outperforming the USA in terms of robots. In 2023, China installed 276,300 industrial robots, six times more than Japan and 7.3 times more than the United States.

There are also a number of domains where the US and China are collectively dominating the global spheres.

Overall, the competing capacities if US and the China is described as the battle between “brains” and “bodies”. The US has traditionally led on so-called AI brains: the world of chatbots, microchips, and large language models (LLMs). China has been superior on AI "bodies": robots, in particular, "humanoid" robots . What if China and the US cooperate, instead of competing? That can result in interesting outcomes. You can read the Nature article for more details.

As the number of indicators above highlights, the so-called “brain” (the US) and “body” (China) are gradually becoming a duopoly in the AI race.

Europe: Too Much Talk, Too Little Action

Europe is not absent from the AI race, but it is weaker than its political language suggests. The European Union has positioned itself as a regulatory power, most visibly through the EU AI Act. Transparency rules, auditability requirements, and compliance frameworks may shape how AI is used, but they, unfortunately, do not create data centers, frontier models, cheap energy, cloud sovereignty, or venture capital depth.

Including the UK, Europe has strong researchers, good universities, and promising firms such as Mistral. Yet it struggles to turn research into globally competitive companies at scale. European startups often move to the United States to access capital, customers, and compute. According to a recent survey, 62% of technology start-ups in Europe run primarily or entirely on US cloud and AI providers. Moreover, one in three European founders, or 29%, said they would consider relocating their headquarters to the US to improve access to capital. No US founders reported considering such a move.

When it comes to skilled labor, many top European AI researchers leave for the U.S. at graduate-school level and remain there.

The infrastructure gap is even more serious. Mistral CEO Arthur Mensch has warned that European dependence on American cloud infrastructure for AI training is a strategic vulnerability. Even when European firms develop strong models, the underlying training infrastructure, chips, and cloud capacity often remain controlled by non-European actors.

Asia’s Hardware Advantage, Europe’s Structural Weakness

Outside the U.S. and China, some Asian economies have found clear access points into the AI economy. South Korea and Taiwan are the most important examples.

The AI chip boom has pushed Taiwan and South Korea ahead of the United Kingdom in global market rankings.

This shift is driven by companies that sit at the heart of AI infrastructure: Taiwan Semiconductor Manufacturing Company, the world’s largest chip foundry, and South Korea’s Samsung Electronics and SK Hynix, which dominate key memory segments. TSMC now ranks among the largest companies in the world, with a market capitalization of around US$1.8 trillion. Samsung and SK Hynix together are valued at around US$1.5 trillion. For comparison, the combined market value of all technology stocks in the Stoxx Europe 600 Index is around US$1.4 trillion.

Asia is benefiting from the hardware foundations of the AI revolution, while Europe remains more focused on regulation, financial services, and fragmented digital initiatives.

This is where middle powers become important. No middle power can realistically replicate the entire AI stack of the United States or China. But they can combine complementary strengths: energy, geography, hardware, talent, industrial applications, and market access.

For instance, electricity prices in Turkey are around one fourth of the European average. Since electricity is one of the most important cost factors for data centers, this positions Turkey as a leading potential location for data-center investment.

Dark Future: New Vassal States

The AI divide is not only an economic problem. It is also a security problem. Today, if the U.S. or China restricted a country’s access to AI systems hosted in their territories, the immediate consequences might still be limited. Most hospitals, military systems, power grids, and public services do not yet depend fully on frontier AI.

But given the picture above, this will change in the near future. As AI becomes embedded in critical infrastructure, defense systems, logistics, finance, health care, and public administration, dependence on foreign AI platforms will become a source of strategic vulnerability. Both the United States and China have already shown in other domains that they are willing to use technological and economic dependence as leverage. There is no reason to assume AI will be different.

This is why data centers, cloud infrastructure, and energy policy are now matters of national strategy. Countries that do not control any meaningful part of the AI infrastructure stack may eventually find their political room for maneuver constrained by those who do.

Governments around the world are beginning to consider AI capability as earlier generations treated oil and electricity grids: as infrastructure on which economic strength, military capability and political independence depend. Some even see computing power as a prerequisite for sovereignty itself, much as nuclear technology defined geopolitical power in the 20th century.

That means, for the so-called “middle powers” there is a huge challenge ahead. Som much concentration of AI power in the United States and China creates a long-term strategic risk the rest of the world, as well as middle powers. Asia, especially South Korea and Taiwan, has found opportunities through hardware and semiconductor ecosystems. Europe, by contrast, risks falling behind because of high energy costs, weak commercialization, fragmented cloud infrastructure, and excessive confidence in regulation.

The gap between the U.S.–China AI axis and the rest of the world is widening. Unless middle powers coordinate/cooperate around energy, compute, hardware, talent, and capital, the AI divide will become a durable structure of technological dependence.

-end-

Recent figures show that just two countries employ 70% of the world’s top machine-learning researchers, command 90% of global computing power, and attract the vast majority of AI investment—more than twice the combined total of all other states.

Dr. Robert Hercock is an experienced AI professional and the Chief Research Scientist for BT’s Security Research Practice. He has more than 20 years of experience managing security research projects in the UK and previously served as theme leader for Networks and Cyber Security at the UK MOD Information Fusion Defence Technology Centre.

Bessemer Converter: Breaking the Energy Bottleneck

To contextualize this tech monopoly, Dr. Hercock drew a powerful historical parallel to 19th-century Sheffield steel production. Before the 1850s, producing high-quality steel via the crucible method was slow, expensive, and heavily constrained by the local availability of timber and coal.

Production was limited by physical energy bottlenecks. That changed in 1856 with the Bessemer Converter. Instead of burning scarce external fuel to bake out impurities, Henry Bessemer introduced a method that blew air through molten iron, allowing the carbon within the iron to act as its own fuel. This breakthrough in efficiency slashed energy costs, bypassed resource limitations, and allowed Sheffield to mass-produce steel on an unprecedented scale.

From Logic Gates to Biological Learning

A similar paradigm shift also played out in the evolution of computing power.

Dr. Hercock traced the history of computing to show how architectural shifts fundamentally redefined what is possible. Early computing research at institutions like MIT in the mid-20th century focused on “symbolic reasoning.” This was a top-down approach where programmers tried to hardcode explicit logical rules into computers. While highly energy-efficient, it ultimately failed because the real world is simply too complex to be captured by rigid logic gates.

The breakthrough came when the they abandoned rigid top-down logic for biological mimicking. Instead of programming explicit rules, pioneers like Geoffrey Hinton (Nobel Prize laureate for his work on artificial neural networks) championed Machine Learning, building bottom-up artificial neural networks inspired by biological brains. By feeding these networks massive datasets, the AI learned to recognize patterns and deduce its own rules through statistical trial and error. This bio-inspired approach contributed the AI revolution, powering the advanced models we see today.

The Modern Resource Bottleneck

Today, however, AI development is hitting a new physical wall. Modern frontier models demand an unsustainable volume of electricity, advanced (and rare) semoconductors and millions of liters of water for data center cooling.

If exponential growth in chip supply continues, AI data centers will need 68 GW in total by 2027 — almost a doubling of global data center power requirements from 2022 and close to California’s 2022 total power capacity of 86 GW.

Under the current paradigm, the US and China's domination relies to a great extend to the fact that they have the capital, infrastructure, and advanced semiconductors to absorb this immense resource footprint. They are treating the AI race as a war of attrition, scaling up power grids and data centers to sustain their lead.

Recalibrating Global Sovereignty

We may be on the verge of a shift toward highly efficient, lean AI architectures that require a fraction of the energy, data, advanced semiconductors and water used today.

If and when these algorithmic breakthroughs occur, the massive infrastructure advantage currently held by the US and China might no longer be an unbeatable advantage. Gigantic data centers will still have value, but they may lose their status as the only way to build advanced AI.

Just as the Bessemer process opened up steel production to the world, a breakthrough in algorithmic efficiency could make AI capability accessible to many more countries. It would allow smaller, agile states to bypass the heaviest resource bottlenecks—not necessarily erasing the lead of tech superpowers, but loosening the US-China grip on the technology and opening up a more balanced future for global power.

It was the hazelnut harvest season.

We were in one of those fully green, distinctive mountain villages of Türkiye’s Black Sea coast. Harvesting hazelnuts means constantly bending down and standing back up—it’s honestly exhausting work. Thank goodness we had our youth; the whole harvest pretty much rests on their shoulders, and they are really good at it.

But something unusual happened: one day, they suddenly dropped what they were doing and headed toward the city center, Samsun.

So, what on earth could make them leave the hazel gardens behind, go to the city center in the dead heat of summer?

An exciting football match? Cinema? A day at the beach?

None of the above.

They were rushing to a massive technology festival in Samsun. And they were far from alone. On that day, more than one million people and over one hundred thousand young innovators developing technology projects gathered in the same place.

I am talking about Teknofest, a phenomenon that has swept across even the most remote corners of Türkiye.

Today, Teknofest is the largest technology festival in the world. At its core, it is both a defense and civilian technology competition. But how did it emerge? How did it suddenly become the coolest thing among young people in Turkiye? And perhaps most importantly, what lessons can other countries learn from this remarkable success story?

Let us take a look behind the scenes.

The View from Oxford

For some time, I have been working as a Visiting Academic at the University of Oxford’s Centre for the Changing Character of War Centre. My research focuses on how civilian technologies and innovative ideas are transferred into military and defense applications. As part of this work, I continuously read, review, and analyze global developments and emerging trends.

Looking at today’s geopolitical landscape, one reality stands out clearly: recent regional conflicts and wars have made domestic defense-industrial ecosystems a necessity rather than a luxury. As a result, countries around the world are searching for ways to strengthen their defence industry base and technological sovereignty.

There are two sides to this challenge. On the one side, as expected, large defence companies, so-called defence primes, deliver major technologies and equipment. On the other hand, governments are trying to identify and scale the small, agile startups capable of producing disruptive technologies and breakthrough innovations. Many countries recognize the importance of this second pillar, yet struggle when it comes to implementation or realization. That means they struggle to attract young people, bringing them together around the challenge and turning the ideas into a product. In fact, that is what we call the technology and innovation ecosystem.

This is precisely where the findings from our fieldwork and academic research on Teknofest offer a potentially transformative model for countries seeking to accelerate innovation in both entrepreneurship and defense technology.

More Than a Festival: What Is Teknofest?

The first Teknofest was held in September 2018 at Istanbul Airport. The mastermind and organizer of the Teknofest is Selcuk Bayraktar, the founder of the famous drone producer - Baykar.

Given the year it started, Teknofest is a relatively young organization. Its mission from the beginning was clear and straightforward: to increase young people’s (aged between 12 and 25) interest in technology and encourage them to develop their own projects, empowering them. Yet from its very first year, it grew rapidly, evolving into a globally recognized aerospace, technology, and innovation festival.

The numbers tell the story:

• The competition spans 47 main categories and 118 subcategories, covering everything from underwater drones and model satellites to vertical-landing rockets and artificial intelligence.

• Each category presents participants with clearly defined technological challenges and real-world problems to solve.

• Young people form teams, develop projects, and submit their ideas.

• Teams that pass the initial evaluation receive substantial funding and technical support to transform their concepts into working prototypes.

Those who successfully advance to the next stage are invited to showcase their prototypes at the festival itself. On the ground, thousands of innovative projects compete for recognition. Above them, Türkiye’s indigenous military aircraft and advanced aviation technologies demonstrate the country’s technological capabilities.

The latest edition attracted 4.6 million visitors, an extraordinary figure that illustrates the scale of the movement.

A New Definition of “Cool”

Teknofest has done something that few public initiatives ever achieve: it has created a genuine cultural shift.

Among young people, the meaning of being “cool” has changed. Participating in Teknofest, working on a challenging technology project, and even advancing far enough to secure prototype funding has become a significant source of prestige.

Local municipalities, youth centers, schools, and universities have embraced this momentum. Some provide workshop facilities. Others assign technical mentors to guide teams through the process. I have personally served as a judge in several competition categories. Reviewing hundreds of projects, I witnessed students from some of the most remote villages and towns in Anatolia tackling highly complex engineering problems with remarkable determination and creativity.

Entrepreneurship and Defense Innovation

But what does the evidence tell us?

Together with two academic colleagues, I conducted a large-scale survey examining a central question: How does Teknofest influence entrepreneurial intentions among young people, and how does it contribute to the emergence of startups that may eventually become drivers of the defense industry?

We are currently preparing the findings for academic publication, but the initial results are striking.

Our research indicates that Teknofest significantly enhances:

1. Opportunity recognition and exposure to entrepreneurial role models, helping participants identify possibilities that others might overlook.

2. Entrepreneurial self-efficacy—the belief that “I can build a company,” “I can develop a product,” and “I can bring an innovation to market.”

Participants begin to view everyday problems through the lens of inventors and entrepreneurs.

From a defense perspective, Teknofest functions as a strategic talent-generation platform. Consider the competition categories: unmanned aerial vehicles, rockets, jet engine designs, air-defense systems, autonomous maritime platforms, and many more. These are not abstract academic exercises (without real-world application); they are technologies directly related to future defense concepts. Many of the winners of the Technofest competition established their own company, by the way.

Thousands of teams spend months designing, building, testing, and refining solutions in precisely these fields.

When it comes to other underpinnings of the Technofest ecosystems, Maker Labs & Deneyap Workshops are the ones exposing young people to technology at an early age across the country. Science and Art Centers (BİLSEMs) & Youth Centers are identifying talented students and connecting them to the innovation ecosystem. Incubators are helping promising competition projects evolve into startups and companies by providing them with physical infrastructure. Many other local institutions assist students and teams.

It is a large-scale societal movement that brings together institutions, universities, companies, and—most importantly—young talent around a shared technological vision.

Today, many countries are searching for formulas to build indigenous defense industries and develop disruptive technologies.

Türkiye’s Teknofest experience offers a compelling model for how a nation can drive mass technological transformation and cultivate the human capital needed for the future.

In a world increasingly defined by technological competition, Teknofest demonstrates that building a defense ecosystem is about inspiring the young generation to imagine, design, and build the future, and turning the whole ecosystem into the new cool.

A 22-point manifesto. Thirty-two million views. A company that named itself after Tolkien’s seeing stones — orbs Tolkien himself described as instruments of corruption — and that now projects itself as the philosopher-king of the West. When Palantir Technologies published its vision for the future of civilisation, it was not offering philosophy. It was making a declaration of dominion and demanding that humanity be grateful for it.

On 18 April 2026, Palantir Technologies posted to its X account under the deceptively casual heading: “Because we get asked a lot.” What followed was a 22-point summary of The Technological Republic: Hard Power, Soft Belief, and the Future of the West, co-authored by Palantir chief executive Alexander Karp and head of corporate affairs Nicholas Zamiska.^[1] Within seventy-two hours, the post had been seen more than 32 million times — the overwhelming majority expressing alarm or outrage. Critics across multiple continents labelled it technofascism.^[3] What it revealed about Palantir — the company behind it, its origins, its contracts and its reach — was more alarming still.

THE ORACLE AND ITS ORIGINS

Peter Thiel named Palantir after the palantíri — the seeing stones of Tolkien’s Middle-earth, orbs of extraordinary power that Tolkien described as instruments of corruption. The irony remains as sharp as ever. Founded in 2003, the company received its crucial early funding from In-Q-Tel, the venture capital arm of the United States intelligence community.^[4]

Karp joined as chief executive in 2004 with an unusual background: a doctorate in neoclassical social theory from Goethe University Frankfurt. Some of scholars have traced the firm’s philosophical genealogy to Thiel and Karp’s shared reading of Leo Strauss — the political philosopher most associated with the idea that elites may be entitled to truths the public cannot handle.^[5] Straussian traces are not difficult to detect in a manifesto that dismisses democratic moral deliberation as “theatrical.”

THE EMPIRE OF CONTRACTS

In fact, Palantir is not a technology company in any ordinary sense. It is a surveillance and data-analysis corporation whose revenues depend upon the perpetuation of precisely the geopolitical conditions its manifesto advocates. Between 2008 and February 2026, the United States Department of Defense awarded Palantir contracts worth more than $2.1 billion.^[6] In 2018, it replaced Google as the contractor for Project Maven, the Pentagon’s AI-powered drone surveillance platform.

In Britain, an investigation by The Nerve identified at least 34 UK state contracts totalling a minimum of £670 million.^[7] The £330 million NHS Federated Data Platform and a £240.6 million Ministry of Defence contract awarded without competitive tender have generated sustained alarm.^[8] The British Medical Association called for “a complete break from Palantir technologies in the NHS.”^[9]

Most consequentially, Palantir’s AI platforms were embedded in Israeli military operations following a January 2024 “strategic partnership” with the Israeli Defence Ministry. UN Special Rapporteur Francesca Albanese found “reasonable grounds” to believe that Palantir supplied the AI platform powering automated target-generation systems that have produced kill lists at a scale no human analyst could match.^[10]

Amnesty International has described Palantir as having “a track record of flagrantly disregarding international law.”^[11] The company operates in thirty-three countries. It describes its own mission as helping to “scare enemies and, on occasion, kill them.”

A DECLARATION OF DOMINION

The manifesto’s ideological content is not incidental. It is structural. The 22 points constitute, point by point, the philosophical justification for precisely the arrangements that make Palantir’s revenues possible.^[1]

Point 1: Silicon Valley owes a “moral debt” to the nation-state. Translation: tech companies should deepen their entanglement with the defence apparatus.

Point 4: Moral reservations about autonomous AI weapons are “theatrical.” Translation: ethical scrutiny of systems like Lavender should be dismissed.

Point 6: National military service should be universal. Translation: a more militarised society is a more profitable one for a defence contractor.

Point 14: The “psychologisation of politics” must be resisted. Translation: any politics shaped by empathy for the vulnerable is an obstacle to civilisational survival.

Point 21: Some cultures are “dysfunctional and regressive.” Translation: the hierarchical ordering of human beings that makes AI targeting systems politically viable now has a philosophical foundation.

Eliot Higgins of Bellingcat was precise: these 22 points are “not philosophy floating in space — they’re the public ideology of a company whose revenue depends on the politics it’s advocating.”^[3] The points amount to an act of corporate self-coronation. Previous concentrations of power — empires, monopolies, aristocracies — generally refrained from publishing philosophical manifestos instructing humanity to be grateful for its own subjugation. This is not the arrogance of a king who believes God granted him dominion. It is the arrogance of an algorithm: systematic, self-referential, and constitutionally incapable of asking whether its premises are correct.

TECHNO-FEUDALISM AND THE NEW LORDS OF DATA

Techno-feudalism is not a metaphor. In the classical feudal order, the serf worked the lord’s land and surrendered a portion of the harvest in exchange for protection. In the techno-feudal order, the citizen inhabits the lord’s platform, surrenders the entirety of their behavioural data, and receives a service whose terms can be altered or weaponised without notice. The platform is the land. The data is the harvest. The manifesto is the declaration of noble right.^[12]

Alongside techno-feudalism, a related concept has emerged: techno-lordism — the self-conception of those at the apex, the belief that because they built the systems on which modern life depends, they are uniquely qualified to determine how modern life shall be permitted to proceed. The Technological Republic is, in its essence, the manifesto of a techno-lord: a declaration that the new feudalism is not merely inevitable but morally necessary, and that those who resist it are obstacles to civilisational survival.

THE DEPENDENCY TRAP

The manifesto’s most insidious proposition is the assumption that undergirds all twenty-two points: that technological dependency is a natural condition rather than a constructed one. It is not. Every government that runs critical functions on Palantir’s platforms; immigration enforcement on Falcon, military targeting through Gotham, patient records on Foundry has transferred a portion of its sovereign capacity to the corporation that controls the underlying system.

Artificial intelligence accelerates this dynamic. When a government’s decision-making depends on AI systems whose training data and inference processes are proprietary and opaque, it surrenders the capacity to understand or contest the outputs that advise its most consequential choices. The deeper the dependency, the greater the power of those who control the system — and the more credible their claim that the relationship should be extended rather than curtailed.

HUMANITY’S COUNTER: AN ENGINEER WHO SAID NO

Against the techno-lordist vision, a response has emerged from within the world of advanced technology itself. Selçuk Bayraktar, chairman of Baykar Technology has spoken directly to the question the manifesto attempts to foreclose: what is the human being’s proper place in an age of machines? A man who builds military drones is the one insisting that human conscience cannot be engineered away. His formulation is striking: “A machine does not suffer, a machine does not feel longing, a machine does not ask ‘why.’” Without that capacity, there remain only “infinite loops, programmed futile tasks, a darkness that has lost its meaning, and ultimately inevitable extinction and absolute destruction.”^[13]

At the SAHA 2026 defence exhibition in Istanbul, Bayraktar broadened the diagnosis. He named the threat: a “techno-capitalist global hegemony” that arrives as “voluntary servitude,” making billions addicted as if to a narcotic. Platform algorithms, he argued, are tuned to exploit neurological vulnerabilities, optimising content around “anger, hedonism and fear” to keep users on the screen for ten more seconds. The smartphone in our pocket has become, in his precise formulation, part of a “spider’s web of digital dependence.” Large technology monopolies now wish to govern the world as “techno-feudalist principalities.” He then named his target directly: a recently published manifesto that sees war, peace and human life as nothing more than “optimisation problems” and “algorithmic outputs”, a text that removes conscience and morality from the equation. Against these “arrogant and dark manifestos,” he argued, those committed to human dignity must place their own roadmap. “Technology in the hands of those without heart, without mercy,” he warned, “can only become a tool of destruction.”^[14]

FROM PALANTÍRI TO TEPEGÖZ: A DIFFERENT ROADMAP

Bayraktar did not stop at diagnosis. Drawing on the ancient Turkish epic of Dede Korkut, he set against the manifesto a counter-image from his civilisation’s mythic memory: the story of Basat and Tepegöz.^[15] Tepegöz was a monstrous cyclops whose enchanted skin neither arrow nor sword could pierce. The hero Basat defeated him not with greater force but with a paradigm shift — striking the monster’s single weakness and killing him with his own enchanted blade.^[14] If Karp’s company took its name from one mythic tradition — the orbs of corruption — Bayraktar drew his answer from another. The defeat of the techno-feudal monster will not come from greater conventional force; it will come from a deliberate reconfiguration of the rules.

The specifics matter. Bayraktar called for a “Technological Solidarity Alliance” among allied and developing nations — built on open-source, auditable ecosystems, federated learning architectures that keep sensitive data within national borders, edge-AI models independent of monopoly clouds, and quantum-resistant encryption.^[14]

Such proposals could be dismissed as aspirational were they not already being acted upon. France’s La Suite Numérique is replacing American software across central ministries.^[16] Germany, Denmark and Austria are following. Switzerland has refused Palantir on national security grounds.^[8] The Tepegöz mythology is not a consolation. It is a strategy.

CONCLUSION: THE REPUBLIC THAT BELONGS TO EVERYONE

The 22 points of The Technological Republic constitute a philosophy of civilisational self-erasure — one in which the human conscience is decommissioned in the name of efficiency, deterrence, and profit. Every point that calls for deeper tech-military integration, for the dismissal of ethical objections, for the ranking of cultures in hierarchies of worth, is a point that makes Palantir’s arrangements more secure and its contracts more lucrative.^[1][6] This is not coincidence. It is the document’s purpose.

A republic, technological or otherwise, is not the property of those who built its infrastructure. It belongs to the human beings who live under its terms. Against the last feudalism, the only adequate counter is a refusal that is principled and organised: a declaration that the future does not belong to those who build the fastest algorithm. It belongs to everyone.

The new lords of data have built themselves an invulnerability much like Tepegöz’s: their power is so embedded in the world’s infrastructure that conventional regulation glances off it as arrows glanced off the monster’s skin. What will defeat them is what defeated Tepegöz — not greater force, but a different architecture, built by those who refused the spider’s web and chose, instead, to ask why. History has seen such convictions of superior right before. It has not tended to remember them kindly.

NOTES & SOURCES

[1] Palantir Technologies, “Because we get asked a lot. The Technological Republic, in brief,” X (formerly Twitter), @palantirtech, 18 April 2026. Over 32 million views within 72 hours. Cited in Fortune, 22 April 2026; Euronews, 22 April 2026.

[2] Cas Mudde, Stanley Wade Shelton UGAF Professor of International Affairs, University of Georgia. Quoted in Common Dreams, “‘One of the Scariest Things I Have Seen’: Alarms Sound Over ‘Technofascist’ Palantir Manifesto,” 21 April 2026. https://www.commondreams.org/news/palantir-technofascist-manifesto-criticism

[3] Eliot Higgins, founder of Bellingcat, post on Bluesky, 19 April 2026. Quoted in Al Jazeera, “Technofacism? Why Palantir’s pro-West ‘manifesto’ has critics alarmed,” 21 April 2026. https://www.aljazeera.com/news/2026/4/21/technofacism-why-palantirs-pro-west-manifesto-has-critics-alarmed

[4] Céline Castets-Renard, Professor and Research Chair on Accountable AI in a Global Context, University of Ottawa, April 2026. https://www.uottawa.ca/about-us/news-all/academic-expertise-technofascism-palantirs-manifesto

[5] Geoff Shullenberger and Moira Weigel on the Thiel–Karp intellectual debt to Leo Strauss. https://outsidertheory.com/the-intellectual-origins-of-surveillance-tech/ and https://www.boundary2.org/2020/07/moira-weigel-palantir-goes-to-the-frankfurt-school/

[6] AFSC Investigate, “Palantir Technologies Inc.,” contract database, updated February 2026. DoD contracts $2.1bn (2008–February 2026); Project Maven; FALCON/ICE. https://investigate.afsc.org/company/palantir

[7] Carole Cadwalladr, The Nerve, February 2026; Novara Media, “What Is Palantir? How a US Spytech Firm Penetrated the British State,” 19 February 2026. https://novaramedia.com/2026/02/19/what-is-palantir-how-a-us-spytech-firm-penetrated-the-british-state/

[8] NHS FDP: £330m, seven-year contract (November 2023). MoD direct award: £240.6m (December 2025). Swiss armed forces’ rejection confirmed in UK Parliament. Hansard, Commons debate, 10 February 2026. https://hansard.parliament.uk/commons/2026-02-10/debates/FECED111-4FA8-4343-BFAF-18D6C13808D0/MinistryOfDefencePalantirContracts

[9] British Medical Association statement, 11 February 2026. Cited in The Lowdown NHS. https://lowdownnhs.info/topics/accountablility/palantir-the-controversy-the-contracts-and-the-campaign/

[10] Francesca Albanese, UN Special Rapporteur, report on Palantir’s role in Lavender, Gospel and Where’s Daddy AI targeting systems. Cited in Truthout, 5 September 2025. https://truthout.org/articles/organizers-are-demanding-palantir-drop-contracts-with-ice-and-israeli-military/

[11] Amnesty International statements, February–April 2026. TRT World, April 2026. https://www.trtworld.com/article/e3c96555543c

[12] Yanis Varoufakis, Technofeudalism: What Killed Capitalism (London: Bodley Head, 2023).

[13] Selçuk Bayraktar, Chairman of Baykar Technology and KÜME Foundation. Cited in Mehmet Çelik, “Humans vs. AI: Is There an Alternative Way?” Daily Sabah, 27 April 2026. https://www.dailysabah.com/opinion/columns/humans-vs-ai-is-there-an-alternative-way

[14] Selçuk Bayraktar, vision address at SAHA 2026, Istanbul, 5 May 2026. Full text: Anadolu Agency, May 2026. https://m.aa.com.tr/tr/gundem/baykar-yonetim-kurulu-baskani-selcuk-bayraktar-teknolojik-dayanisma-ittifaki-kurmaliyiz/3932435. All direct quotations from the SAHA address are drawn from this source.

[15] On the Tepegöz legend: Kitab-ı Dede Korkut, eighth chapter (“Basat’ın Tepegöz’ü Öldürdüğü Boy”). UNESCO Intangible Cultural Heritage, 2018.

[16] DINUM, “Souveraineté numérique : l’État accélère la réduction de ses dépendances extra-européennes,” 8 April 2026. https://www.numerique.gouv.fr/sinformer/espace-presse/souverainete-numerique-reduction-dependances-extra-europeennes/

How did the Strait of Hormuz become a “knot”—a choke point capable of paralyzing not only a global military superpower like the United States but the entire world?

While the recent closure of the Strait during the conflict between US/Israeli forces and Iran has devastated global markets, the real story isn’t just economic—it’s geographical. To understand why the world’s largest military power is “cornered” here, we must look beneath the surface.

Shallow Seas and Narrow Corridors

Most of the deep areas of the Persian Gulf are inside Iranian territorial waters. The shallower waters in the southern reaches of the Gulf are strewn with numerous small coral islands, oil-well heads, pumping platforms, and underwater mounds, all of which inhibit navigation and force traffic into designated deepwater channels near Iran’s islands or coastline.

The element of surprise is particularly important in unconventional naval warfare, and is particularly likely in littoral waters because of the relatively small distances involved, and because close-range engagements reduce some of the technological advantages enjoyed by larger forces such as the U.S. Navy. Therefore, Iran is likely to use terrain, camouflage, concealment, and deception measures (including platforms or weapons incorporating low-observable technology and features) to achieve surprise.

At its narrowest point, the Strait is 39 km wide—roughly 6 km wider than the Strait of Dover. However, the “Hormuz Knot” isn’t about width; it’s about bathymetry (underwater topography).

For the VLCCs (Very Large Crude Carriers) that drive global trade, the navigable section is dangerously limited. These ships have drafts of 20–25 meters, forcing them away from the undulating, shallow relief of the Iranian coast and into narrow, deep-water corridors.

• The Highway Effect: Maritime traffic must follow strict inbound and outbound lanes, each only 3.7 km wide.

• The Sitting Duck Scenario: This 11.1 km total “safe zone” forces warships and aircraft carriers to execute sharp maneuvers directly under the shadow of the Iranian landmass. In this narrow corridor, massive naval vessels become stationary targets for land-based missile batteries.

To prevent collisions, specific transit corridors act as “underwater highways.” Ships must use inbound and outbound lanes that are each only 3.7 km (2 nautical miles) wide. Including the buffer zone, this creates a narrow 11.1 km lane that is the only safe route—not just for tankers, but for US aircraft carriers and warships.

The Result: This route heads deep toward the Iranian coast before executing a sharp westward maneuver. In this “turn,” massive naval vessels become sitting ducks. They are forced to slow down and move cautiously, virtually surrounded by Iranian landmasses equipped with high-precision missile batteries.

The Hidden Defenses: Sand Waves and Mud Volcanoes

Iran’s long shoreline is overlooked by a mountain ridge that rises as high as 6,500 feet (2,000 m) and continues uninterrupted along the northern shores of the Gulf. Moreover, the Gulf ’s network of islands, inlets, and coves, and its coastal support infrastructure (buoys, onshore hide sites and bunkers, coastal observation posts), are ideal for staging and supporting extended patrol and reconnaissance operations, precision mining operations, missile barrages, and swarming attacks.

To reopen Hormuz, the US would need to control the Iranian coastline. But geography has deployed its own “invisible” soldiers beneath the waves:

1. The Salinity Engine: The extreme heat of the Persian Gulf creates high evaporation, making its water dense and highly saline. Conversely, the Indian Ocean is cooler and fresher. This creates a massive “conveyor belt” where dense Gulf water flows out along the bottom with incredible force.

2. Sand Waves: This powerful bottom current shapes the seabed into massive current-induced sand dunes and “waves,” making the floor unstable and unpredictable.

3. Mud Volcanoes: At the collision point of the Arabian and Eurasian Plates, tectonic pressure forces gases like methane and CO2 upward, dragging sediment with them to create “mud cones.”

This shallow, uneven, and shifting seabed is the perfect environment for GPS-controlled, mobile naval mines. Because these systems are mobile rather than free-floating, detecting and destroying them with standard mine-hunting vessels is an exercise in futility. The sea is, for all intents and purposes, closed.

A Country Surrounded by Natural Walls

If the sea is impassable, an airborne operation might seem like the quickest solution. Here, however, another geographical wall emerges: High Karstic Topography.

Iran is a fortress. The Zagros Mountains stretch from the Turkish border to the Strait, reaching heights of 4,400 meters. This is Karst topography—hard, calcareous rock riddled with natural caves and labyrinthine tunnels. These conditions are perfect for concealing tactical equipment and mobile launchers. Even if US paratroopers successfully landed, holding ground in this “Swiss cheese” terrain without heavy support would lead to catastrophic casualties.

The entire country is fortified by these “ramparts”:

• West: The Zagros Mountains.

• North: The Alborz Mountains (5,610m).

• South: The Dasht-e Kavir and Dasht-e Lut deserts.

• East: The Makran Mountains.

The Final Verdict: Geography’s Victory

The “Hormuz Knot”, as illustrated, is not merely a military puzzle; it is a geographic reality that defies conventional power. While the U.S. and Israel possess unparalleled technological supremacy, they are effectively “fighting the Earth” in a theater where every mountain, current, and sand wave serves as an Iranian ally.

About the Author

Onur Kovaci is a PhD candidate at King’s College London. He specialises in regional geography, border studies, and geopolitics, with a focus on Cyprus and the Gulf region.

By Mohammed Alharthi

At around 9:45 a.m. Iran time on February 28, 2026, U.S. and Israeli forces opened Operation Epic Fury and Operation Roaring Lion with nearly 900 strikes in 12 hours across Tehran and Iran’s southern military corridor. The theory of victory in the minds of American operation planners was that the war would result in achieving objectives, through perhaps later negotiations or otherwise, that the negotiations in Oman in February 2026 had failed to achieve. But today, as the war marks its 60th day, with a fragile ceasefire in place, a failed postwar negotiation attempt, and a recent American announcement that it will not send a delegation to Islamabad, that theory has yet to materialize.

Despite a widespread regional hope for a return to a stable order and elimination of future threats to it, the war, as of now, has successfully put new strains to the track of the Iran-US negotiations.

Trust

Questioning the credibility of Iranian interlocutors is neither new to the US, nor its regional partners. The nature of the Iranian regime, whether in its oscillating reformist-principlist orientations or the IRGC-presidency internal divergences, has always put a question mark on the positions of Iranian diplomats. Yet, the question of legitimacy has never been more relevant than today. The incompatibility of Iranian operational conduct in the war and statements made by officials signal a mosaic defense strategy following the death of the Iranian supreme leader, Ali Khamenei. There have also been reports of internal challenges to Mohammad-Bagher Ghalibaf, the head of the recent Iranian delegation to the Islamabad talks. The Americans’ confidence in the Iranian shaking signature is, at best, weak.

Yet, recent reports suggest that the war has gradually unified the Iranian leadership. While there is not sufficient data to verify a rally ‘round the flag effect among the population, the emerging leadership in Iran is reportedly closing ranks. This can shadow a step towards legitimacy of Iranian interlocutors. While this unification skews positions towards the IRGC’s noncompromising tendency, the Iranian behavior does respond to incentives; specifically it is survival-sensitive.

From the Iranian perspective, the US also has an issue of interlocutors. The fact that the United States and Israel went to war shortly after a round of prewar negotiations, even if partly due to Iranian intransigence, deepened distrust in the viability of negotiations. President Trump’s rhetoric during the war and recent negotiations can further exacerbate trust. This provides a possible explanation to the Iranian keenness to negotiate with Vice President JD Vance, who is reportedly more anti-war than President Trump.

Trustbuilding

Success of negotiations, especially at the scale that this context presents, requires a serious process of trust building. Both sides have to feel reassured enough that not only are they getting each other’s good faith concessions but also that those concessions remain credible in implementation. The last nuclear agreement between Iran and the United States, the JCPOA, took more than two years to reach, even though it left out issues Washington and its partners considered central, including Iran’s missile program and support for regional proxies. Any new agreement would have to carry a heavier burden. It would likely need to address those previously excluded issues, alongside Iran’s stockpile of 60 percent enriched uranium and the reopening of the Strait of Hormuz. That makes the diplomatic task before the United States and Iran far harder than the one that produced the JCPOA. One way to expedite an essential trustbuilding process, reduce distrust, is through guarantors and mediators who can make commitments more credible.

Without a resourceful guarantor, the negotiations are Hamlet without the Prince. Pakistan now seems to be the most immediate assumer of the role. It has access to Tehran to be taken seriously, while still having —notably recent—working history with Washington. Its distance from the Gulf arena, albeit hedged by a thick defence alliance with Saudi Arabia, also gives it some comfort to the Iranians.

The harder question is whether Pakistan alone can carry the weight that a guarantee in this negotiation would require. Any guarantor would have to make each side believe that the other side’s commitments can survive the next change in political mood. Pakistan may help open the door. It is less clear that it has the leverage of scale to hold the door open on its own.

Another path to double down guarantees, specifically on the American bargain, relies on the Iranians ability to win back the Arab gulf states that continue to pursue a de-escalatory track. While the Iranians’ calculus, in attacking the Gulf states, was to exert pressure on the global economy and particularly the US administration, the gulf has shown resolve in not folding to this play. Today, the gulf states still hold considerable leverage with the American administration. They also have a direct stake at ensuring any agreement is credible and sustainable. Channels of diplomacy remain on the table, but it requires sincere Iranian effort to assure the Gulf of its commitment to a political solution that ensures stability in the region. This needs to involve concessions that the IRGC must view as a price well worth paying for the future of Iran.

Indeed, Iran seems to have embarked on this journey of regionalizing the negotiations process. Yet, it is still far away from the required concessions.

Another missing piece to the guarantor puzzle is China. Although China does not have the most credible track record given that one of its most recent, and clearest, guarantee commitment was undermined by this very war. Whether or not the Chinese have an impressive resume for the guarantor job does not undermine the fact that they are the country who hold the most leverage to guarantee the actions of the Iranian regime. The Chinese may deem the guarantor role too expensive as it gives up the flexibility it currently enjoys in the region. Whether it uses its leverage or not must be a critical point to address by the remaining mediators and the US.

Leverage

Prior to the war, the Americans focused on three issues vis a vis Iran. These three were the nuclear program (enrichment and enriched uranium), missile and drone capabilities, and support for regional proxies. But the war added the opening of the strait of Hormuz, and guarantee of the strait’s safety in the future, as an additional timely demand. On the other side, faced with the pressures of an unbalanced war, the Iranians have not only a need to relieve sanctions but also to ensure it no longer is threatened with coercion. The latter can possibly manifest in a non-aggression agreement or a maintenance of deterrence tools.

The Americans continue to hold the sanction card. For Iran, sanctions relief is getting more urgent with time. The Iranian social contract, like many in the region, requires a certain level of state provision to maintain manageable calm among the population. Especially with the American decision to stop Iranian shipping through a strait already closed by the Iranians, the regime’s financial sustainability is nothing short of challenging.

Time is not necessarily on the American side, however. With the closure of the strait of Humouz, Iran has created a new leverage in this war. The Trump administration also faces its own domestic pressures, as potentially manifested through loss of GOP congressional seats in November of this year, and spillovers to the presidential elections in 2028. President Trump is also bound by a reactive stock market and limited munitions to regional allies.

Between the pressures of time, the Iranian pursuit of keeping tools for regime survivability, and the American desire to prove their theory of victory—that the war created a situation that enables the US to achieve objectives they couldn’t retain otherwise—the negotiations, and indeed the war, will continue to be a game of attrition. This game is characterized by an asymmetry of risk tolerance in favor of Iran and an overwhelming balance of power, albeit hedged by Iranian insurgency tactics, in favor of the US.

To find a way out of the attrition trap, one strategy likely to be attempted is phasing negotiations to allow negotiators to take contentious items one step at-a-time. On one hand, the Americans are very likely to insist on the nuclear-related demands as a requirement of the first phase, and the Iranians will try to limit the scope at the strait of Hurmouz closure. Some of the other demands can serve as a swing, getting the negotiations to the finish line. But the Iranians’ readiness to give up these outright is low. This makes another round of fighting more likely.

Prospects

Both the American and Iranians have a choice to make between trust-building or remaining on their track of testing each other’s resolve. Without proper trust-building, the negotiation prospects are not most favorable. States interested in a war-ending agreement will continue seeking the critical mass of guarantors. Otherwise, additional American attempts to degrade Iranian capabilities is only more likely— sooner or later.

Biography of the Author

Mohammed Alharthi studies the tools of foreign policy and statecraft. He is a returning associate at McKinsey & Company, where he works on elevating governments’ performance. He has served in the United Nations’ Departments of Political & Peacebuilding Affairs and Peace Operations and chaired a UN interns committee that channeled youth priorities with UN Ambassadors. At UCLA, he earned a B.A. in Political Science and a B.S. in Mathematics & Economics with highest honors, completing an award-winning thesis on the statecraft of Gulf States using LLMs. His early experience includes working in strategy at the Saudi Industrial Development Fund and advancing expertise at UCLA’s Center for Middle East Development. Mohammed consults California-based startups on Gulf markets in AI, robotics, and Industry 4.0. Engaged in youth affairs, he co-founded Furas to build an internship ecosystem in KSA. He sits on the Youth Program Advisory Committee of Misk Global Forum and the General Assembly of Nuqsh. At Oxford, Mohammed is completing the Master’s of Public Policy at the Blavatnik School of Government.

Just a few years ago, defence cooperation between Rome and Ankara was viewed primarily as a technical element of NATO’s southern flank. Today, given the changes taking place within the alliance itself, it increasingly resembles a strategic political project that could have far-reaching consequences, especially for the future of the European security architecture.

The paradox of current relations, however, is that they are taking place at a time when the European Union is being more vocal than it has been in years about „strategic autonomy,” building a common arms industry, and the need for greater independence from the United States. Despite this rhetoric, the actual military capabilities Europe urgently needs—especially in the fields of drones, air defence systems, and unmanned technology—are increasingly coming not from Germany, France, or Spain, but from Türkiye, which, despite years of efforts, is still not a formal member of the European Union. In this sense, Italy’s turn toward Türkiye is not just a pragmatic economic choice, but also a signal that some European countries (though not all) are involuntarily beginning to see Ankara as an informal, though increasingly indispensable, partner in Europe’s security.

Why Türkiye?

From an Italian perspective, Türkiye possesses several characteristics that make it an exceptionally attractive partner.

First, Ankara has the second-largest army in NATO, after the United States. Its armed forces operate in the Black Sea, Syria, Libya, Iraq, East Africa, and the Eastern Mediterranean. Türkiye is therefore not only a weapons manufacturer but also a country with operational and combat experience, further enhanced by its key geographic location.

Second, the Turkish defence industry is developing faster than most of its European competitors. Over the past decade, Ankara has developed its own drone, missile, naval, and air force programs. Turkish companies are able to offer equipment faster, more cheaply, and often with a greater willingness to transfer technology than their Western competitors.

Third, Türkiye remains a key player in regions crucial to Italian security policy: Libya, North Africa, the Middle East, and the Eastern Mediterranean. Rome realises that without cooperation with Ankara, it will be unable to effectively influence issues such as migration, energy security, or the stabilisation of Libya.

In practice, this means that Türkiye is beginning to play a role vis-à-vis Italy that the EU cannot provide, for example, by supplying ready-made military capabilities and enabling rapid action when European programs remain delayed or ineffective.

At the same time, this cooperation is gaining importance not because the two countries share identical interests, but because there is a gap between Europe’s growing security needs and the limited willingness of the European Union – specifically, some member states hostile to Türkiye – to include Turks in new defence initiatives. As a result, Rome and Ankara are developing cooperation primarily on a bilateral basis and within NATO.

SAMP/T and air defence: Turkey enters Europe through Italy

One of the most symbolic examples of this new rapprochement is the talks regarding the SAMP/T air defence system.

Türkiye has been pursuing an advanced long-range air and missile defence system for over a decade. Ankara has been interested in the SAMP/T system since 2014 and, in 2018 signed a potential co-production agreement with the Eurosam consortium.

However, this project was put on hold, primarily due to French resistance. As a result, Türkiye opted to purchase Russian S-400 systems, leading to a crisis in relations with the United States and Ankara’s exclusion from the F-35 Lightning II program.

Today, the situation seems to be different. Events in recent months – particularly the tensions surrounding the US-Israeli-Iran conflict and the related missile incidents near NATO member states’ borders – have shown Türkiye that it needs a modern air defence system sooner than previously anticipated. According to reports, four missiles were intercepted over or near Turkish airspace by allied systems. Ankara saw this as evidence of the urgent need to strengthen the country’s air defence.

It is worth noting, however, that Türkiye is still reluctant to buy ready-made products, as this involves the risk of increasing maintenance and repair costs in the future and, consequently, making the country dependent on a foreign partner. Ankara is therefore setting very specific conditions: technology transfer, participation in production, localisation of part of the industrial process, and access to know-how. This is precisely why Italy is becoming so important to Türkiye. Rome is seen as a more flexible partner than, for example, Paris, and Italian companies can act as intermediaries between Turkish needs and the European defence industry.

If an agreement is reached on the SAMP/T system, its significance will extend far beyond air defence alone. It will mean that Italy will become Türkiye’s main European partner in the field of strategic technologies.

Bayraktar TB3: Italy turns to Turkish drone power

An even more spectacular example of this convergence is the Italian Navy’s planned purchase of Bayraktar TB3 drones for the aircraft carrier Cavour.

The Italian Navy’s decision is historic, as it will make Italy the first European operator of an armed drone capable of operating from the deck of an aircraft carrier. At the same time, it will be the first time an EU country has decided to integrate Turkish technology so deeply into its own armed forces.

The Bayraktar TB3 represents the next stage in the evolution of the TB2 platform, which is also used by Poland and gained international recognition through its performance in conflicts in Ukraine, Libya, and Syria. The TB3 version has been adapted to operate from the short decks of aircraft carriers and amphibious assault ships. The drone features foldable wings, reinforced landing gear, and the ability to carry precision weapons. Crucially, the TB3 is ready „here and now,” while European drone programs are still in the design, discussion, or chronic delay stages. For example, the Eurodrone program faces significant financial and political challenges. The United Kingdom has tested MQ-9B Mojave drones but has not yet made a procurement decision. France, meanwhile, is considering withdrawing from some unmanned systems projects.

Türkiye, on the other hand, has a finished product that has already been tested in combat and operational conditions.

It is also worth noting that Italy’s decision to purchase the TB3 follows the drone’s highly successful performance during Steadfast Dart 2026 in the Baltic Sea, the Alliance’s largest exercise of the year. Approximately 10,000 troops from 13 countries participated in the exercise, notably without the participation of US forces. In practice, the exercise also became a test of Europe’s ability to project force independently without direct US support.

In this context, the TB3 performed exceptionally well. Three prototypes operating from TCG Anadolu completed over 200 combat sorties from January to March 2026, operating in extremely challenging weather conditions, with temperatures dropping to minus 6 degrees Celsius, strong winds, and heavy snowfall. The highlight, however, came on February 14, when the drone autonomously launched from a ship, destroyed a naval target with two precision-guided missiles, and then safely returned to the deck. This was the first full combat cycle of a carrier-based drone in the history of NATO exercises – from takeoff, through attack, to landing. The TB3 also completed an eight-hour, 1,700-kilometer flight alongside Eurofighter Typhoon fighters, demonstrating the interoperability of manned and unmanned platforms.

For the Italian Navy, this was a sufficient demonstration, convincing Rome that Türkiye had not only developed a modern unmanned system but had also built a new naval doctrine around it. Italy’s purchase of the TB3 therefore represents more than just filling a technological gap. It is also a political recognition that Türkiye has surpassed Europe in some areas.

Leonardo and Baykar: the birth of a new industrial model

The collaboration between Leonardo and Baykar occupies a unique position in the process of building international cooperation in the defence sector, demonstrating that such a partnership does not need to rely solely on the export of finished products. The Italian and Turkish companies have created a joint industrial structure – LBA Systems – dedicated to the design, production, and maintenance of unmanned systems, which appears to meet the needs and expectations of both parties. For Italy, this means access to drone technologies that are still limited in Europe. For Türkiye, it means entering the European market through a reputable EU partner.

Furthermore, it offers a way to circumvent political constraints. Products jointly developed by Leonardo and Baykar can be perceived as „European”, even if the key technologies originate in Türkiye. This allows Ankara to access certain programs and funds that are formally closed to non-EU countries.

Maritime cooperation as a new strategic frontier

Italian-Turkish cooperation is not limited to drones. Maritime technologies are also gaining increasing importance.

In this context, it is worth mentioning the companies Havelsan, Piloda Defense, and VN Maritime Technologies, which are jointly developing new unmanned and hybrid surface vessels for Italian state services. While the equipment will be manufactured in Italy, the autonomous technologies, management systems, and artificial intelligence will come from Türkiye.

It is worth noting that these platforms will be used not only by the Navy but also by the Italian Coast Guard, the Guardia di Finanza, and institutions responsible for the protection of the marine environment. This means that Turkish technologies can be integrated into the daily functioning of the Italian state in the areas of internal security, surveillance, and border control, demonstrating that Ankara is becoming not only an arms supplier but also a full-fledged industrial partner.

Italy prioritises pragmatism over ideology

It cannot be ignored that Ankara continues to arouse distrust in many European capitals, primarily because of its relations with Russia, the purchase of S-400 missiles, its policy toward the Kurds, and its actions in Syria and Libya. Rome is not indifferent to these issues either and does not seek to ignore them. However, from the perspective of political interests, pragmatism appears to have recently prevailed in Italy. Italian elites are gradually reaching the conclusion that Europe can no longer afford to marginalise Türkiye. Ankara possesses the military potential Europe needs, i.e., combat experience, industrial capacity, and a strategic geographic position.

Photo: Turkish National Defence Minister Yaşar Güler met with Italian Defence Minister Guido Crosetto. Photo. Ministry of National Defence of the Republic of Türkiye - official website

This also leads to a significant change. For years, EU relations with Türkiye were defined primarily by political disputes, human rights issues, and the frozen accession process. Today, they are increasingly shaped by specific strategic interests such as drones, missiles, warships, air defence, and security in the Mediterranean.

Italy has traditionally pursued a policy based on three „circles”: Atlantic, European, and Mediterranean. In practice, this means that Rome simultaneously strives to maintain strong relations with the United States, strengthen its position in the EU, and play a stabilising role in the Mediterranean. However, after 2022, Italy’s priorities have clearly shifted. The war in Ukraine forced Rome to become more involved on NATO’s eastern flank, while maintaining its traditional southern focus. For Italy, Türkiye is therefore becoming not so much a problem as a necessity, a fact well known to the government led by Giorgia Meloni, who differs significantly from her predecessor, Mario Draghi. The current Italian Prime Minister places greater emphasis on national sovereignty, border security, migration, the protection of critical infrastructure, and cooperation with conservative partners.

Italy, therefore, needs Türkiye as a regional partner and intermediary in areas where it has limited influence. Türkiye, in turn, needs Italy as a gateway to European political, industrial, and defence structures.

At the same time, it should not be forgotten that Turkish-Italian industrial cooperation does not eliminate rivalry. Türkiye is increasingly competing with Italy in export markets, particularly in Africa, the Middle East, and Central and Eastern Europe. The Turkish arms industry is developing faster, benefits from lower labour costs, and is more willing to transfer technology than European companies. This could limit the competitive advantage of Italian manufacturers in the long term.

Conclusions

The Italian-Turkish rapprochement poses a fundamental question for Europe: should Türkiye be part of the European security architecture, or will it remain merely an external partner used on an ad hoc basis?

For now, the EU’s answer remains ambiguous. Formally, Türkiye remains outside most EU defence programs. In practice, however, European countries are increasingly utilising Turkish potential.

This creates a growing contradiction between rhetoric and reality. Europe speaks of strategic autonomy but is unable to independently deliver many of the necessary technologies. As a result, it must rely on Turkish solutions.

Italy seems to understand this reality more quickly than others. Rome is not waiting for a common position from Brussels, Berlin, or Paris. Instead, it is building its own model of cooperation with Ankara, which is based on pragmatism, co-production, and shared strategic interests.

The conference has emerged as one of the most prominent security forums in the wake of the Ukraine War, first convened in 2023. Bringing together policymakers, military officials, industry leaders, academics, and media representatives, the conference provides a multidimensional platform to debate the rapidly evolving global security environment.

Each year the LDC has focused on distinct themes, with 2023’s “Stronger Security in a Fast-Changing World,” 2024’s “Deterrence: Building Capacity to Counter Global Threats,” and 2025’s “Alliances” reflecting shifts in international security. This year the conference held between April 9-11, centered on the theme of “Readiness.” This year’s theme evokes a stronger sense of action than the previous ones. Rather than merely diagnosing threats or discussing frameworks, “Readiness” calls for proactive preparation, rapid adaptation, and the ability to respond effectively to a wide spectrum of security scenarios.

The opening panel “Uniformed Services and the Whole-of-Society Approach” set the tone by highlighting the necessity of integrating civilian and military efforts in national defence strategies. This approach recognizes that modern conflicts include traditional battlefields but also extend beyond them, encompassing economic resilience, technological innovation, and societal cohesion. In this sense, one of the key concepts discussed extensively during the conference was resilience. Unlike traditional notions of strength, resilience is a core component of readiness and relates to the ability to learn, adapt, and sustain functionality under pressure.

Following the invasion of Ukraine, several European countries have invested in enhancing resilience in civilian domains. Some have prepared public guidance materials outlining actions to take in times of crisis or war, while others have explored ways to protect cultural heritage during conflicts. Discussions at the conference reflected a growing interest in strengthening the whole-of-society approach in the UK. The challenge lies in framing the narrative of readiness in a way that mobilizes society by fostering a sense of urgency without inducing fear.

Additionally, the concept of reindustrialization emerged as a central subtheme of the conference, not merely as an increase in production capacity but as a qualitative shift toward smarter, faster, cheaper and more collaborative manufacturing processes. During the sessions, it was emphasized that the nature of warfare is undergoing rapid transformation, driven by technological advancements and shifting geopolitical dynamics. National security now encompasses a wide array of domains, from supply chains and energy systems to information environments and emerging technologies. Defence industries are no longer evaluated solely on output volume but on their ability to innovate, integrate, and scale efficiently in response to emerging threats. In this sense, readiness is not a static condition but a dynamic process shaped by continuous learning and adaptation.

The conference also addressed the evolving character of warfare, with particular attention to lessons learned from Ukraine. The conflict has demonstrated the continued relevance of conventional warfare while simultaneously highlighting the transformative impact of new technologies. The widespread use of drones has fundamentally altered the battlefield, enabling more flexible and cost-effective operations. One of the most important examples of Ukraine’s advanced capabilities in this regard is Operation Spider Web, a drone attack targeting Russian aircraft which was the subject of my research poster presented at the conference. In this study, I analyzed the extent to which it corresponds to existing theories of deception and what it reveals about the continuity and transformation of deception in modern warfare. Operation Spider Web is one of the most significant examples showing how weaker states can gain asymmetrical advantages over stronger adversaries through deception and surprise attacks. Ukraine’s experience further demonstrates how necessity can drive rapid innovation, offering valuable insights for other countries seeking to enhance their defensive capabilities.

The concept of alliances remained central to the conference’s discourse. For the UK, while NATO continues to be the primary framework for collective defence, strengthening bilateral and regional partnerships within Europe and through the “special relationship” between the UK and the US is essential. The idea that “united we are stronger” encapsulated the prevailing sentiment, especially in the face of a perceived alignment among adversarial states such as China, Russia, Iran, and North Korea. Despite their ideological differences, these countries were described as converging around a shared objective of challenging the Western-led international order. A recurring concern was the perceived uncertainty surrounding the US’ commitment to European security and its role within NATO.

Ultimately, the LDC underscored a fundamental shift in how security is conceptualized and practiced. Readiness is not a static goal but an ongoing process that demands continuous adaptation, collaboration, and investment. It requires a holistic approach that integrates military, economic, technological, and societal dimensions, while also addressing the psychological aspects of preparedness.

Looking ahead, the continuation of initiatives such as the Future Leaders’ Day and the upcoming the LDC Washington Forum in September 2026 indicates that the conference aims to sustain its momentum and expand its influence. As global security challenges become increasingly complex and interconnected, platforms like the LDC will play a crucial role in shaping Western strategic thinking and fostering the collaborative efforts needed to navigate an uncertain future.

Spring has arrived in Oxford with unusual speed this year. Locals say it has been one of the rainiest seasons in memory. Yet with the arrival of blossom and longer days, nature has come alive – and so has our planning for an even more productive period ahead. We at the Oxford Centre for Geopolitics have taken this moment to reflect, recalibrate, and plan for a productive and ambitious period ahead.

So, what exactly have we been doing over the past semester?

Over the past semester, our focus has been on laying the intellectual and institutional groundwork for the year ahead. As global geopolitical dynamics continue to evolve at pace, the need for fresh perspectives and rigorous analysis has never been more pressing.

Our 2026 Research Agenda

This year, the Oxford Centre for Geopolitics will focus on the following three core themes:

1. Future of AI: Geopolitics and the Global Divide
   Exploring how artificial intelligence is reshaping global power structures and widening gaps between technological leaders and followers.

2. When Civilian Technology Becomes Strategic Power: The New Era of Dual-Use Innovation
   Examining the increasing overlap between civilian and military technologies, and its implications for global competition.

3. Shifting Power: China, Middle Powers & the West
   Analysing the evolving balance of power between major and emerging actors in the international system.

These topics reflect the most pressing intersections between technology, power transitions, and international order. Detailed research papers, policy briefs, and events will be rolled out throughout the year.

Launch of Our New Web Platform

We are pleased to announce the launch of our new website ( https://www.oxford-geopolitics.org/ ), alongside a dedicated publishing platform on Substack - https://opinion.oxford-geopolitics.org/. This platform will serve as the primary outlet for our fellows’ work, featuring weekly analysis and commentary on pressing geopolitical issues.

We warmly invite you to subscribe and follow our publications as we expand our intellectual footprint.

Trinity Term Activities

Looking ahead to Trinity Term, we have an exciting lineup of events and engagements planned:

• A series of seminars and roundtable discussions hosted in Oxford University

• Visits to leading think tanks in London

• A forthcoming collaborative event with King’s College London and Cambridge University (details to be announced)

• Two joint events with the Oxford International Development Centre

• Additional events at Saïd Business School and Pembroke College

• A collaborative event with the Oxford Space Initiative

Further details will be shared soon—stay tuned.

Welcoming Our New Fellows

We are thrilled to welcome the 2026 cohort of Research Fellows to the Oxford Centre for Geopolitics. This year’s fellows hail from leading institutions including the University of Oxford, University of Cambridge, King’s College London, and the University of Warwick.

These distinguished scholars bring fresh expertise and diverse perspectives. Over the coming months, they will share their research and insights through articles and briefings on our new Substack platform.

We wish them every success and are confident their contributions will greatly enrich our work and the wider policy debate.

As Trinity Term begins and Oxford fills with the energy of spring, the entire team at the Oxford Centre for Geopolitics is energised and ready for what promises to be a dynamic and impactful period.

We look forward to keeping you informed and engaged throughout the term.

On April 8, 2026, France became the first major Western democracy to formally commit its entire government apparatus to abandoning Microsoft Windows in favour of Linux. The announcement, made at an interministerial seminar convened in Paris at the Prime Minister’s direction, is less a technology procurement decision than a geopolitical one: a direct response to escalating fears that American software infrastructure has become a strategic liability for European states. DINUM — the Direction interministérielle du numérique, France’s coordinating body for government IT — will migrate its own workstations as an initial proof of concept. Every ministry and public operator must then formalise its own dependency-reduction plan by autumn 2026, covering seven domains: operating systems, collaborative tools, antivirus software, artificial intelligence, databases, virtualisation, and network and telecommunications equipment.¹

The ambition is sweeping. France is attempting something no large democracy has accomplished at national scale — systematically unwinding decades of dependence on a single foreign technology ecosystem.

THE TRIGGER WAS POLITICAL, NOT ONLY TECHNICAL

The timing of the announcement is no accident. Since January 2025, the Trump administration has wielded technology access as a coercive instrument in ways that have alarmed European capitals. When the United States sanctioned International Criminal Court officials, ICC chief prosecutor Karim Khan reportedly lost access to his Microsoft email account — an incident Microsoft subsequently disputed, claiming the ICC itself made the decision to disconnect his services, though Dutch press reporting suggested Microsoft had made the broader service continuation contingent on that disconnection.² US tariffs of 20 per cent on EU goods, Vice-President Vance’s attacks on European regulation at the Munich Security Conference, and Commerce Secretary Lutnick’s explicit linkage of tariff relief to the rollback of digital rules have collectively dissolved any remaining illusion of apolitical technology partnerships.

David Amiel, France’s Minister of Public Action and Accounts, framed the pivot in blunt terms: “We can no longer accept that our data, our infrastructure, and our strategic decisions depend on solutions whose rules, pricing, evolution, and risks we do not control.” His counterpart Anne Le Hénanff, Minister Delegate for Artificial Intelligence and Digital Technology, described digital sovereignty as “a strategic necessity,” positioning France as the vanguard of a broader European shift.³

The legal architecture reinforces the political anxiety. The US CLOUD Act of 2018 grants American authorities the power to compel any US-controlled company to produce data stored anywhere in the world — including servers physically located in the European Union. FISA Section 702, reauthorised in 2024, enables mass surveillance of non-US citizens’ data without individualised warrants. These statutes exist in irreconcilable tension with Europe’s GDPR. In June 2025, testifying before a French Senate inquiry into public procurement and digital sovereignty, Microsoft France’s director of public and legal affairs acknowledged that, should US authorities issue a legally justified demand, the company could not guarantee that data held in France would not be transmitted without the explicit consent of French authorities.⁴ With American cloud providers controlling the vast majority of the European cloud market, the exposure is not theoretical.

A SOVEREIGN SOFTWARE STACK ALREADY TAKING SHAPE

France’s bet on Linux is not a leap into the unknown. It builds upon a two-decade precedent that most outside observers have overlooked. The National Gendarmerie — France’s 95,000-strong military-police force — began its migration from Windows to a custom Ubuntu-based distribution called GendBuntu in 2004, taking a deliberately phased approach: first replacing Microsoft Office with OpenOffice, then migrating browsers and email clients, and finally deploying Linux on desktops from 2008 onwards. By June 2024, 103,164 Gendarmerie workstations — representing 97 per cent of the entire fleet — were running GendBuntu, constituting one of the largest Linux desktop deployments in any government worldwide.⁵

The Linux migration is the most dramatic element of a broader campaign to replace American software across French government operations. DINUM has assembled La Suite Numérique — a cohesive ecosystem of open-source productivity tools already adopted by more than 500,000 users across more than 15 ministries and affiliated public bodies.⁶ The suite encompasses four functional pillars.

Tchap, a secure messaging platform built on the open Matrix protocol, has exceeded 375,000 monthly active users and was made mandatory across government ministries from September 2025, with its use explicitly preferred over platforms such as WhatsApp, Signal, and Telegram for official communications. Visio, a video conferencing tool built on the open-source LiveKit framework with AI transcription provided by French start-up Pyannote, is replacing Microsoft Teams and Zoom: the CNRS — France’s national research centre, which is planning the replacement of Zoom licences covering 34,000 agents and 120,000 affiliated researchers — has been among the institutions preparing migration. France Transfert, a sovereign file-sharing service hosted on ANSSI-certified SecNumCloud infrastructure operated by Dassault Systèmes’ subsidiary Outscale, has displaced WeTransfer and Dropbox for government file exchange. The National Health Insurance Fund (CNAM), with 80,000 agents, has formalised a partnership with DINUM to deploy La Suite tools — and France’s national health data platform, currently hosted on Microsoft Azure, is scheduled for migration to sovereign infrastructure by the end of 2026.⁷

Digital sovereignty requires more than swapping desktop software — it demands indigenous capacity in artificial intelligence, where dependence on American platforms may prove even more consequential than operating-system choices. Here, France possesses a genuine industrial asset in Mistral AI. Founded in 2023 by alumni of DeepMind and Meta, the Paris-based company reached a valuation of €11.7 billion in its September 2025 Series C, led by Dutch chip designer ASML, with Nvidia, Andreessen Horowitz, Bpifrance, General Catalyst, Index Ventures, and Lightspeed as additional participants. ⁸

France’s national AI strategy, initially launched in 2018 and substantially expanded in 2023, has aligned public policy with the thesis that sovereign computing infrastructure — not merely regulatory frameworks — is required for meaningful AI independence. Mistral CEO Arthur Mensch has argued publicly that dependency on American cloud infrastructure for AI training constitutes a strategic vulnerability, and that European organisations require alternatives hosted within European jurisdictions. Supported by both private investment and state policy alignment, Mistral is building the physical compute layer that France’s sovereignty ambitions require.

EUROPE IS MOVING — BUT THE SCALE GAP REMAINS FORMIDABLE

France is not acting in isolation. Germany’s Schleswig-Holstein has migrated roughly 80 per cent of its 30,000 government workstations away from Microsoft, reporting annual licensing savings of €15 million in 2026 alone. Denmark’s Ministry of Digital Affairs announced its own open-source transition in 2025. Austria’s armed forces have replaced Microsoft Office with LibreOffice across 16,000 computers. The European Data Protection Supervisor found in 2024 that the European Commission’s own use of Microsoft 365 violated GDPR. ⁹

The EU AI Act, reaching full application in August 2026, adds regulatory momentum. Its requirements for transparency and auditability in high-risk AI systems structurally favour open-source architectures — an advantage for European providers such as Mistral, whose models are substantially open-source, over their closed American competitors. The convergence of the AI Act, the Data Act applicable since September 2025, DORA for financial services, and NIS2 for critical infrastructure is creating an environment in which sovereign technology choices become not merely politically attractive but legally prudent. ¹⁰

Yet the scale challenge is formidable. American hyperscalers — AWS, Microsoft Azure, and Google Cloud — collectively invest hundreds of billions annually in infrastructure, dwarfing all European efforts combined. GAIA-X, the Franco-German federated cloud initiative launched with considerable fanfare in 2019, has largely failed to produce operational infrastructure commensurate with its political ambitions. European cloud providers hold only a small fraction of their own continental market, a share that is declining rather than growing. Closing the infrastructure gap would require sustained, coordinated EU investment on a scale that is nowhere on the political horizon.

DIGITAL SOVEREIGNTY: DATA, DEFENSE AND DESTINY?

France’s digital sovereignty agenda faces a paradox familiar to all long-term technology transitions: the timeline for execution extends well beyond the political cycle that generated the mandate. The presidential election looms approximately one year away. Microsoft’s European headquarters sits in Paris; lobbying pressure will be intense and sustained.

The deeper tension is structural. Even as France replaces its desktop layer, the substrate beneath remains overwhelmingly American. Government AI models train on NVIDIA hardware. Sovereign aspiration is not the same as sovereign reality.

If the autumn 2026 ministry plans prove genuinely ambitious, and if the June 2026 Industrial Digital Meetings generate durable public-private coalitions, France will have created an institutional framework capable of outlasting any single administration. The Gendarmerie’s quiet, two-decade Linux deployment survived multiple changes of government precisely because it delivered measurable operational results without ideological grandstanding.

Whether a national-scale transition can replicate that pragmatic persistence — amid tariff wars, artificial intelligence races, and electoral cycles — will determine whether France’s digital sovereignty remains coherent policy or retreats into mere political posture. The borders of the 21st-century state, France is insisting, are no longer drawn only on maps. They are written in code. And Paris intends to hold the pen.

NOTES & SOURCES

¹ DINUM (Direction interministérielle du numérique). “Souveraineté numérique : l’État accélère la réduction de ses dépendances extra-européennes.” Official press release, 8 April 2026. https://www.numerique.gouv.fr/sinformer/espace-presse/souverainete-numerique-reduction-dependances-extra-europeennes/

² The Register. “Microsoft throws spox under the bus in ICC email flap.” 18 February 2026. Multiple reports confirmed Karim Khan’s email access was disconnected following US sanctions; Microsoft disputed direct responsibility. https://www.theregister.com/2026/02/18/microsoft_asks_uk_parliament_to_correct_record/

³ Heise Online. “France’s plan: Away from Windows, towards Linux.” Quotes from ministers Amiel and Le Hénanff at the April 8 seminar.https://www.heise.de/en/news/France-s-plan-Away-from-Windows-towards-Linux-11251739.html , 9 April 2026.

⁴ Forbes. Microsoft France Director Anton Carniaux, French Senate public procurement inquiry hearing, June 2025. Carniaux stated: “No, I cannot guarantee that.” https://www.forbes.com/sites/emmawoollacott/2025/07/22/microsoft-cant-keep-eu-data-safe-from-us-authorities/

⁵ The Next Web. “France orders all government ministries to ditch Windows for Linux in digital sovereignty push.” GendBuntu statistics confirmed: 103,164 workstations, 97% of fleet, €2M/year savings, 40% TCO reduction. https://thenextweb.com/news/france-linux-windows-migration-digital-sovereignty

⁶ DINUM / numerique.gouv.fr. LaSuite partnership announcement with CNAM: “already adopted by more than 500,000 users” and deployed across “more than 15 ministries.” https://lasuite.numerique.gouv.fr, March 2026.

⁷ LeMagIT. “Polémique autour de LaSuite.” February 2026. Tchap: 375,000 monthly active users; mandatory from September 2025. CNRS: 34,000 agents and 120,000 researchers. Outscale/Dassault SecNumCloud hosting confirmed. https://www.lemagit.fr/actualites/366638848/Polemique-autour-de-LaSuite-la-DINUM-se-defend-et-ne-ferme-pas-la-porte-au-prive

⁸ ASML / Mistral AI. Series C funding press releases, 9 September 2025. €1.7 billion raised; €11.7 billion post-money valuation; ASML lead investor with €1.3 billion. https://www.asml.com/en/news/press-releases/2025/asml-mistral-ai-enter-strategic-partnership ; https://mistral.ai/news/mistral-ai-raises-1-7-b-to-accelerate-technological-progress-with-ai

⁹ Schleswig-Holstein migration: https://interoperable-europe.ec.europa.eu/collection/open-source-observatory-osor/news/schleswig-holsteins-open-source-strategy-year

¹⁰ European Commission. AI Act (2024); Data Act (applicable September 2025); NIS2 Directive; DORA.https://www.eiopa.europa.eu/digital-operational-resilience-act-dora_en

“In the race to master artificial intelligence, the United States and China are not just building faster algorithms; they are constructing rival blueprints for the digital world’s future.”

Share

The US-China race in artificial intelligence is about more than algorithms and processors: it is a struggle to control the rules, the infrastructure, and the future of the digital world.

Just as the nuclear arms race became the strategic centrepiece of the Cold War, the contest for dominance in artificial intelligence (AI) and quantum computing (QC) is fast becoming the defining arena of twenty-first-century great-power rivalry.

These capabilities are no longer peripheral research projects. They are now central to national security, poised to transform economic productivity, military effectiveness, and geopolitical influence.

The country that establishes and secures a decisive lead will not merely set the pace for the next generation of civilian and military applications; it will have the means to recalibrate the global balance of power.

Because AI and QC are inherently dual-use technologies, advances in one area can quickly be repurposed in the other. This enables the development of more autonomous weapons systems, more sophisticated surveillance architectures, and faster, data-driven strategic decision-making.

Neither Washington nor Beijing believes it can afford to lose this race.

For Washington, conceding leadership is not an option. The prevailing consensus is that the United States must remain ahead to safeguard both economic advantage and national security.

Reflecting this urgency, Vice President JD Vance described the development of artificial intelligence as an ‘arms race’ with China.

Donald Trump’s return to the White House has ushered in a more assertive approach to AI policy driven by the ‘America First’ agenda.

Early in his second term, he issued an executive order titled ‘Removing Barriers to American Leadership in Artificial Intelligence’, explicitly committing his administration to accelerating the development of AI with minimal constraints as a means of ensuring US technological dominance.

Shared goal, divergent paths

If AI is to become the foundation of global power, the question is no longer “whether” states will compete for dominance, but “how”—and on whose terms.

Both Washington and Beijing are convinced that mastery of AI infrastructure will determine who sets global standards, secures the world’s data flows, and controls the computational resources underpinning finance, defence, and beyond.

But they diverge in their methods, principles, and narratives.

This logic is strikingly reminiscent of the nuclear competition of the Cold War, when fears of a ‘missile gap’ with the Soviet Union led to an intense acceleration of US strategic programmes.

Then, as now, policymakers saw technological victory as the key to sustaining global primacy.

For Washington, allowing a rival, particularly China, to control this domain would mean relinquishing strategic leverage over the networks and platforms that define the modern world.

Its newly unveiled “Winning the AI Race: America’s AI Action Plan” opens with an unequivocal statement of intent: “The United States is in a race to achieve global dominance in artificial intelligence (AI).

Whoever has the largest AI ecosystem will set global AI standards and reap broad economic and military benefits.”

The plan’s America First orientation casts AI leadership as a core pillar of US national security.

It proposes dismantling domestic regulatory constraints, accelerating innovation, expanding data-centre capacity, and ensuring semiconductor self-sufficiency— measures designed to safeguard America’s proprietary advantage and maintain its geopolitical leverage in the AI era.

This approach focuses on controlled access, involving the restriction of advanced computing and chipmaking technology exports to China and the sharing of cutting-edge capabilities primarily with ‘trusted allies’.

This exclusivity-focused approach stems from the belief that protecting proprietary advantages is essential for maintaining commercial leverage and strategic superiority.

China’s ‘openness’ strategy

Beijing responded to Washington’s action plan with one of its own, emphasising international cooperation in both development and regulation.

Warning against AI becoming an “exclusive game” controlled by a few nations and corporations, China positioned itself as the champion of openness.

China’s Global AI Governance Initiative has pledged to expand open-source development and share capabilities with developing countries, particularly in the Global South, as well as to promote multilateral governance through a new international body headquartered in Shanghai.

These initiatives are intended not only to institutionalise China’s leadership role but also to bring global rules, standards, and norms more closely into alignment with its strategic interests.

Two objectives underpin this openness.

First, by releasing advanced open-source models, Beijing is contesting the US approach of monetising AI as an exclusive, proprietary product. If these open-source platforms achieve the same level of performance as their US counterparts, the commercial advantage of exclusivity — one of Washington’s key competitive assets — could be significantly diminished.

Second, by enabling these models to be freely adapted and exported, China embeds its technology in the digital infrastructure of emerging economies. This creates long-term dependencies and gives Beijing greater leverage over global technical standards and governance frameworks.

Who will dominate AI sovereignty?

Recent advances by Chinese AI firms are closing the performance gap with the US, raising the prospect of a world in which American dominance is no longer guaranteed — and in which Chinese models could be adopted on a par with, or even more widely than, those in the US.

In such a scenario, the contest would shift from technical superiority to control over the standards, ecosystems, and governance models embedded in global digital infrastructure.

As Microsoft President Brad Smith warned, “The No. 1 factor that will define whether the US or China wins this race is whose technology is most broadly adopted in the rest of the world. Whoever gets there first will be difficult to supplant.”

Beijing’s centralised, state-led model enables it to mobilise resources, align industry and exploit disruptive opportunities far more quickly than the US’s private-sector–driven approach. This agility could put Washington in a paradoxical position: technologically advanced, yet strategically outmanoeuvred.

For middle powers and emerging economies, the stakes are no less acute. Both Washington and Beijing export not just products but technological ecosystems that carry political influence.

Alignment with either superpower risks long-term dependency; avoiding both would require costly investment in domestic capacity—a politically difficult and time-consuming path.

Washington’s AI Action Plan implicitly presents its ready-made solutions as the pragmatic option. However, history shows that dependence on external providers often comes with strategic constraints.

Therefore, the concept of Sovereign AI—the capacity to develop, deploy, and govern a nation’s own AI systems without relying on foreign infrastructure—is gaining prominence.

Sovereign AI is not simply about technological pride; it is about ensuring control over where AI workloads run, how critical data is stored and used, and who ultimately has the power to switch those systems on or off.

In sensitive sectors—defence, intelligence, energy, and critical infrastructure—such autonomy could act as a strategic shield, insulating states from coercion, espionage, or supply-chain disruption.

Many countries are more tempted by the immediate option of choosing turnkey systems offered by Washington’s proprietary stacks or Beijing’s open-source platforms.

However, both options pose risks. US solutions often tie users into closed commercial ecosystems, while Chinese offerings, although open in form, can establish deep structural dependencies that are hard to undo.

The real question is not simply which superpower will “win” the AI race, but whether the rest of the world can muster the capacity—and the political will—to carve an independent path.

Otherwise, sovereignty in the digital era may become the first casualty of great-power competition.

This article was originally published by TRT World on August 15, 2025. The views expressed herein are solely those of the author and do not necessarily reflect the views or editorial stance of TRT World.

Subscribe now

“As Europe’s biggest economy moves beyond pacifism on paper, it remains trapped between the habits of the past and the demands of a more dangerous world.”

Share

Russia’s 2022 offensive in Ukraine represented the most significant break in European security since World War II, signalling a decisive departure from the post-Cold War order.

While Europe has endured numerous crises over the years, none has so fundamentally shattered the continent’s assumptions of peace and stability.   

The incursion not only breached the long-standing principle of inviolable borders but also brought back the spectre of a major war into Europe’s heartland. 

The illusion of protection from large-scale conflict—once widely believed across much of the continent—collapsed suddenly, prompting a reassessment of national security doctrines.  

In response, states that had long depended on the protective umbrella of others are now facing a harsh new reality, leading to urgent efforts to rebuild their defensive capabilities. The change is not just reactive; it signifies a deeper reconsideration of strategic autonomy and resilience.   

Uncertainty over US security commitments, China’s assertiveness, energy dependencies, increasing cyber vulnerabilities, and intensifying political fragmentation within the EU have all heightened Europe’s strategic unease.   

No country has experienced this rupture more viscerally than Germany. 

For over thirty years, Berlin built its prosperity on an export-driven model powered by cheap Russian gas, while leaving its deterrence to NATO and allowing the Bundeswehr to weaken.

That delicate balance fell apart in 2022. 

Germany now faces an existential threat. The ontological foundations of its post-war identity—as a civilian power with a deep aversion to militarisation—have become increasingly untenable. 

Reconstructing that identity requires more than mere adaptation. It demands a deliberate break with the past—and a willingness to forge a new role in a much less forgiving world.  

Otherwise, if existential anxiety remains unresolved, Germany faces paralysis: a state caught between normative ideals and strategic needs, unable to act decisively in an era that punishes hesitation. 

Material commitments and strategic ambiguity

In the aftermath of Russia’s offensive in Ukraine, Germany has undertaken its most ambitious military overhaul since WWII. The extent of investment, institutional reform, and symbolic acts is notable.   

Yet behind the surge in spending and political rhetoric, a fundamental question remains: is this a genuine strategic reorientation—or a series of reactive measures under pressure?   

For decades, Berlin’s defence stance relied on two closely linked pillars: a civilian power role based on diplomacy, restraint, and legalism; and a geo-economic power identity rooted in the belief in the civilising force of commerce.

And nowhere was this more evident than in its relationship with Russia.

Within this framework, affordable Russian gas was more than just an economic resource—it upheld Germany’s self-image as a responsible player within a stable international order. 

This dual role, although coherent in theory, posed a strategic contradiction in practice. Germany appeared as a reluctant power—hesitant to project force, cautious of leadership, and slow to make commitments.

Paradoxically, this very reluctance offered a form of ontological security: it allowed Germany to reconcile its NATO commitments with ongoing economic relations with Moscow.

However, the binary confrontation triggered by the Russian military attacks in Ukraine shattered that balance.   

The consequences have been extensive. Then-chancellor Olaf Scholz’s Zeitenwende speech in February 2022 marked a turning point, quickly followed by a constitutionally protected €100 billion fund to modernise the Bundeswehr.   

Germany is now on track to increase its annual defence budget to €162 billion ($191B) by 2029—a 70 per cent rise from 2025 levels—potentially making it Europe’s largest military spender. 

Core defence outlays are projected to reach 3.5 per cent of GDP, significantly above NATO’s 2 percent benchmark.   

To finance this shift, Chancellor Friedrich Merz has introduced constitutional reforms that relax Germany’s long-standing fiscal constraint known as the debt brake.

Up to €1 trillion ($1.2T) in borrowing has been authorised over the next decade for defence and infrastructure, breaking with decades of debt aversion. A €380 billion ($448.4B) package is already designated for military modernisation through 2029.   

Equally notable is Berlin’s effort to break cultural taboos surrounding military power. These taboos are deeply rooted in Germany’s Nazi past.

The post-1945 'never again' ethos turned anti-militarism into a defining feature of national identity, viewing the use of force not just as a strategic decision, but also as a significant moral risk.

In 2024, Germany held its first official veterans’ commemoration since World War II. A long-overdue national security councilwas created, fulfilling a campaign pledge by Merz.  

And in a historic first, Germany started its first permanent foreign deployment to Lithuania—marking the first overseas peacetime deployment of its forces since WWII. 

Yet despite the momentum, doubts remain. Critics claim that key decisions—such as cutting Russian gas, sending Leopard tanks to Ukraine, and purchasing F-35s—were influenced more by external pressure than by strategic vision.

Germany’s defence procurement system continues to be slow and inefficient, raising concerns about its capacity to effectively implement projects.   

And despite the rhetoric of transformation, it remains unclear whether the ‘Zeitenwende’ signifies a fundamental shift in strategic thinking—or merely a necessary response to a deteriorating status quo. 

The social limits of strategy

Germany’s post-2022 defence transformation has been politically bold and materially expansive. But while budgets have increased and military plans have accelerated, the strategic culture supporting this change remains underdeveloped.  

The ‘Zeitenwende’—now more than two years old—has unfolded less as a coherent national strategy and more as a series of elite-driven, reactive responses to external shocks. What is still missing is a fundamental alignment between military ambition and societal consent.  

The clearest sign of this gap is in the manpower issue.

While the Bundeswehr plans to grow from 182,000 to 260,000 soldiers by the mid-2030s, and reserve forces are to increase from 60,000 to 200,000, planners face a harsh reality: a population that is increasingly disengaged from the armed forces.   

The 2011 abolition of conscription under former chancellor Merkel was not just a policy change—it reflected a cultural consensus that the military should remain peripheral to German civic life. That consensus has not yet been meaningfully reversed.  

The newly approved voluntary military service programme aims to recruit up to 40,000 young people each year by 2031 and will require all 18-year-old men to undergo medical assessments—even those not planning to enlist. However, early signs suggest resistance rather than enthusiasm.   

Recent polling indicates that only a third of Germans aged 18–29 support reinstating conscription. Among younger Germans, concerns about a potential draft are increasing—evidenced by a rising interest in how to legally avoid service.  

This recruitment hesitancy is worsened by the government’s failure to reconnect with former conscripts because of outdated record-keeping and data protection laws. Defence Minister Boris Pistorius has already stated that if the volunteer model fails, conscription will be reinstated.  

This is more than a personnel dilemma—it is a warning sign about the limits of transformation without participation. A better-equipped Bundeswehr will not be a more credible deterrent if it lacks social legitimacy.  

Defence expenditure is expected to reach €162 billion ($191B) by 2029, and Germany is progressing towards meeting NATO’s new target of dedicating 5 percent of GDP to defence.

However, these financial commitments alone cannot fill the strategic void created by public indifference and political division. 

The implication is clear: without a societal ‘Zeitenwende’—a broad recalibration of how Germans understand risk, power, and responsibility—material improvements risk reinforcing, not resolving, Germany’s status as a reluctant power.  

A state can announce brigades and purchase F-35s, but if it cannot convince its own citizens of the legitimacy of force, it remains strategically incomplete. Germany’s challenge is not simply to spend—but to engage. Not just to arm—but to align.  

Without a corresponding shift in strategic culture and societal consent, increased funding and high-profile procurements will simply produce a better-equipped version of the same reluctant power, not a truly independent and responsible actor.

This article was originally published by TRT World on September 15, 2025. The views expressed herein are solely those of the author and do not necessarily reflect the views or editorial stance of TRT World.

Subscribe