NestAI of Finland has raised €100 million and formed a strategic partnership with Nokia to develop artificial intelligence for defence-grade autonomous systems, spanning from unmanned vehicles to command-and-control software. The deal positions the Helsinki, Finland, start-up in the heart of Europe’s quest to develop homegrown “physical AI” — systems that can function under real-world constraints and resist deliberate interference.
Inside the NestAI funding round and Nokia AI pact details
Finland’s state-owned investor Tesi led the round, with Nokia also participating, which suggests a patriotic interest in dual-use technology that could be adopted by allies and the Finnish Defence Forces.
The capital will be used to finance the region’s “leading physical AI lab” in Europe, hiring across robotics and edge inference, as well as accelerating trials of autonomy stacks for land, air, and maritime. In dollar terms, about $110–$115 million; the raise is a notable outlier in an obdurately tight venture market and signals changing risk appetites for defense tech in Europe.
If you take into account historical loss rates, the number of casualties and the barrel cost, “the wine has just got to an extraordinary level” for management teams that have consistently delivered value so far at above-average reinvestment rates, says New Fortress co-founder Richard Clarey. Sarlin, who sold Silo AI to AMD for $665 million, has invested through his family office and will continue as a non-executive as NestAI hires a CEO. Working in stealth mode until now, the company already counts the Finnish Defence Forces as a customer for its AI adoption and has hired from Intel, Kongsberg, Palantir, and Saab.
What Physical AI Could Mean for the Battlefield
Physical AI is a term used to refer to models, which are often large language or multimodal systems, as embedded in machines that perceive the world at a time-sensitive scale, decide and act upon it. In defense, that translates to onboard perception, navigation, and mission logic that keeps drones flying under GPS jamming or lets ground vehicles handle rough terrain, or combines sensor streams into meaningful options for commanders without ongoing human micromanagement.
Doing so requires more than simply training a model in the cloud. Physical AI needs to run at the edge within power, size, and thermal constraints, gracefully recover from degraded communications, and pass strict safety/validation gates. Industry standards like vision-language-action systems out of top labs and NVIDIA’s own robotics toolchains point the way forward, but defense adds adversarial settings, electronic warfare, and export controls to the engineering checklist.
Why Nokia is a complementary fit for defense autonomy networks
Nokia will bring hard-earned credibility in secure, mission-critical connectivity — private 4.9G/5G, edge computing, and resilient networks. For autonomous systems, the communication layer is not a nice-to-have: low and predictable latency, network slicing-based prioritization, and guaranteed links in the presence of interference could be the make-or-break factors if autonomy does more good than harm. Nokia also has hundreds of private wireless deployments in industrial facilities around the world, knowledge that has direct application to highly mobile, sensor-heavy defense settings.
Finland’s rich radio engineering talent pool adds another layer — from legacy assets like Nokia Bell Labs to 6G research hubs located in Oulu. In combination with Nokia’s edge platforms, NestAI’s autonomy stacks could enable on-vehicle inference with single-digit millisecond round-trip times, while also ensuring a secure data backbone for command-and-control. That alignment reduces the risk of integration, a major factor in defense customers’ procurement decisions.
The European defense context shaping autonomy investments
Defence budgets in Europe have increased sharply since Russia’s full-scale invasion of Ukraine. In 2023, military spending in Europe grew at a double-digit rate, its highest since the end of the Cold War, according to SIPRI. Procurement priorities have moved in the direction of air defense, long-range fires, and autonomous systems that can be fielded more quickly than traditional platforms.
Policy tailwinds are strengthening. The European Defence Fund is putting in some €8 billion until 2027 for cooperative R&D, while the NATO Innovation Fund at €1 billion has a remit: dual-use deep tech across member states. European incumbents and startups, from Saab to Helsing, are scrambling to provide AI-driven situational awareness, electronic warfare, and autonomy that satisfy the demands of sovereignty. NestAI’s Finland-first stance is in line with that agenda: control of data, models, and supply chains locally; exportability without exposure to non-European regimes.
Team, governance and the road ahead for NestAI and partners
The staffing of NestAI indicates an intention to straddle the divide between silicon, software, and systems integration — vets from chip design houses and defense primes working under a governance structure that has Sarlin’s chairman role distinct from the day-to-day business.
The company will require strong compliance with export controls and clear human-on-the-loop doctrines to meet European ethical and legal sensitivities, especially as the EU AI Act carves out military use but influences adjacent civil deployments.
“Short-term steps to take include field trials with Finnish units, integration tests over private 5G, and demonstrations of measurable gains — fewer operator interventions per mission hour, higher target detection rates under jamming, smaller logistics footprint in edge inference. Also, interoperability with NATO’s standards for data formats and communications is going to be another gating factor.”
The headline to take away is simple: when combined with a large domestic capital pool and world-class connectivity partner, NestAI is putting Finland on the map as a serious hub for defense-grade autonomy in Europe. If the company can turn its lab promise into independent, certifiable systems, that €100 million bet could pay out with the kind of sovereign AI stack European militaries are growing increasingly interested in — and increasingly in need of.