Table of contents
In December 2025, a tonne of agricultural urea cost American farmers the equivalent of 75 bushels of maize. On 9 March 2026, after US-Israeli raids on Iran had effectively closed the Strait of Hormuz, it cost 126. A 77% jump in less than three months, according to an industry estimate cited by the CSIS. A third of the world’s maritime trade in fertilisers passes through that strait, including 49% of the agricultural urea exported globally. There are no international strategic reserves of ammonia comparable to oil reserves. Nor is there even an established alternative route for fertilisers: the pipeline that Saudi Arabia built to bypass the strait is designed for crude oil, not nitrogen-based products. When we talk about food security as critical infrastructure, this is the figure to start from. Not a hypothetical scenario: a news story from this week.
The geopolitical landscape that agritech cannot ignore
The Hormuz crisis marks the third agricultural input crisis in four years. The first, in 2021–22, was caused by Russian gas: following the invasion of Ukraine, prices for natural gas – a raw material used in the synthesis of ammonia via the Haber-Bosch process – tripled, dragging nitrogen fertilisers up with them. In that instance, farmers were able to offset the costs to some extent. The blockade of the Black Sea grain corridors also drove up wheat prices, and grain producers recouped some of their margins. This time is different. Iran is not a major wheat producer. Grain prices remain low. Input costs are rising without a corresponding rise in the value of the finished product, on already squeezed margins: in the US, agricultural bankruptcies (Chapter 12) rose by 46% between 2024 and 2025, from 216 to 315 cases.
The nature of the problem is not new, but it continues to surprise those who manage agri-food supply chains. The Persian Gulf accounts for almost half of global urea exports and 30% of ammonia exports. Europe has historically depended on imports of nitrogen fertilisers: dependence on Russia and Belarus was estimated at around 20% of global trade, as became starkly apparent in 2022. The supplier changes, but the structure remains the same. Critical inputs for food production pass through unguarded geopolitical bottlenecks, with no purpose-built logistical alternatives and no strategic system reserves.
Anyone developing agrifoodtech in Europe in 2026 is operating within this system, whether they realise it or not
The concept of dual use in agrifoodtech operates on three levels that are rarely discussed together. The first is the classic military-to-civilian transfer: drones, hyperspectral satellite imaging, biosensors and GNSS all originated or developed within defence and intelligence contexts, and today enable precision farming, plant protection and food safety. The second, less obvious level is the reverse. The same technology that detects Fusarium in a wheat field identifies biological agents in a scenario of deliberate contamination. The biosensor monitoring a food production line is the same platform that DARPA funded under the FS2 (Foundational Security for Food Systems) programme, with the stated aim of protecting cereal crops as a national security priority. The third aspect is regulatory. Legislative Decree 138 of 4 September 2024, which came into force on 16 October 2024 and through which Italy transposed the NIS2 Directive, explicitly includes ‘food production, processing and distribution’ among critical sectors, with obligations regarding risk assessment, incident reporting and business continuity management.
Who is already working on it, and with what funding?
The most instructive example is Constellr (Munich). The company was founded to offer precision farming services based on satellite thermal imaging: monitoring crop water stress, soil temperatures and optimising irrigation. On 10 February 2026, following a €37 million Series A funding round (bringing total funding to €75 million), it announced a pivot towards government, military and security agency clients. Same technology, same satellite, same data. Radically different buyers. CEO Max Gulde put it in no uncertain terms: “National resilience requires understanding what assets are actually doing on the ground, not just seeing them.” From agritech to defence, without changing a single photon.
Orbital Sidekick (San Francisco) has followed a similar path with hyperspectral imaging. In 2020, it secured a $16 million contract from AFVentures, the investment arm of the US Air Force, to integrate its SIGMA platform into the Advanced Battle Management System. At the same time, In-Q-Tel, the Intelligence Community’s fund, participated in a $10 million funding round. The same technology used to map crop stress is employed to identify hidden production sites, detect missile fuel and track the movements of non-cooperative assets. This is no mere curiosity: it is a market structure that determines who secures contracts and on what terms.
In the field of biosensors, the link is almost structural. BioFire Defense (Salt Lake City) produces both a panel for biological threat agents and a panel for monitoring pathogens in food and water. Oxford Nanopore Technologies (Oxford) applies portable sequencing to both food outbreaks and the identification of biological threats. DARPA has made this rationale explicit with the FS2 programme: “ensuring the active defence of cereal crops is a critical national security priority”, states the official text. The food safety testing market is currently worth around $26 billion, with forecasts of €48 billion by 2033. A growing share of that growth is driven by technologies whose origins in patents and procurement lie in the defence sector.
When it comes to GNSS, this is perhaps the most significant issue in structural terms. Almost all new European tractors equipped with GNSS use EGNOS for precision steering (source: EUSPA). EGNOS, in its current form, amplifies GPS signals. GPS is managed by the US Space Force. In other words: the autosteer system on almost every European tractor is operationally dependent on an American military system. The transition to EGNOS v3 will integrate Galileo, reducing this dependence, but this will take time. We must also clear up a misconception circulating within the Italian sector: ‘autonomous’ tractors in Europe are still, for the vast majority, GPS autosteer systems. They are not machines that operate without an operator on board. By the end of 2024, globally, there are estimated to be around 65,000 autonomous or semi-autonomous tractors in operation, with the European share concentrated on pilot schemes in large German, French and Dutch farms. AGCO launched its OutRun retrofit kits only in 2025, whilst John Deere unveiled the fully autonomous 9RX at CES 2025, but for the American market. Those constructing narratives about autonomous farming in Europe are describing a near future, not a reality on the ground.
The risks that nobody wants to talk about
The first is dependence on unmanned infrastructure. The Strait of Hormuz has shown that a geopolitical crisis affects agricultural costs within weeks, not years. The same applies to technological dependence: an attack or disruption to GNSS signals in a context of escalation – a scenario already documented in Ukraine and the Baltic Sea – has immediate repercussions on the operation of digitally guided agricultural machinery. GPS jamming and spoofing are techniques already in use in real-world conflict scenarios. No European precision farming system has a business continuity plan for a scenario involving signal degradation.
The second issue concerns data ownership. Satellite imagery, water stress maps, and historical yield data are largely classified as non-personal data, falling outside the scope of standard GDPR protections. Who controls this data, and for what purposes, remains a question that the Italian agri-food sector has not yet asked itself. In a 2026 report, IPES-Food highlighted the risk of technological lock-in and concentration within digital agricultural supply chains. This is not an ideological stance: it is a description of how markets are structured where information asymmetry is total and where data generated in the field ends up being exploited elsewhere.
The third issue relates to compliance, and directly affects start-ups. If a product falls within the dual-use category – and it is often not immediately obvious – export authorisations, restrictions on technical assistance and contractual complexities come into play, which many founders only discover when it is already too late. The cost of compliance is significant and is rarely factored into the go-to-market plans of those building agritech businesses in Europe.
And what about Italy? A position yet to be established
Italy has real strengths in this area. The Italian Space Agency’s PRISMA mission carries a hyperspectral sensor that is also used for crop monitoring. e-GEOS, part of the national space/EO industrial sector, promotes precision agriculture platforms integrated with COSMO-SkyMed. CIRA is part of the DIANA test centre network, the NATO programme for dual-use technologies, with an Italian node of the accelerator called DualTech by Takeoff. There is also a highly instructive precedent in industrial policy. In 2021, the government led by Mario Draghi used its ‘golden power’ to block the acquisition of seed producer Verisem by Syngenta, which is owned by the Chinese giant ChemChina. It was the first time the ‘golden power’ had been exercised in the Italian agri-food sector. The Council of State upheld the veto in January 2023. Seeds, plant genetics, agricultural data: already subject to national security screening today, even if no one puts it in those terms.
What is missing is the link between these assets and the agrifoodtech startup ecosystem. The NATO Innovation Fund manages over a billion euros earmarked for deep tech in areas including the resilience of critical infrastructure. DIANA offers early-stage investment and access to test centres. These are open channels. How many Italian agrifoodtech start-ups have systematically explored them? Very few, because ‘dual use’ is not yet part of the vocabulary of Italian agritech founders, and because accelerator programmes and sector-specific funds do not list them among viable funding options.
Meanwhile, the Strait of Hormuz is closed, fertilisers cost 77% more than they did three months ago, and the spring sowing season won’t wait for diplomats to agree on reopening the sea lanes. Agriculture has become software-defined, gas-dependent and satellite-reliant. Understanding who we depend on, and making an informed choice about who we want to depend on, is no longer merely an exercise in academic geopolitics.
It is a business decision that needs to be taken now, before the next crisis makes it unavoidable.
Note to reader: the author is CEO of Beeco and works with investment funds active in agritech, which may have supported or will support some of the startups mentioned in the future.
ALL RIGHTS RESERVED ©