This week’s defence AI landscape is defined by a dramatic scaling-up of investment and operational deployment: Europe is pouring record capital into sovereign AI defence firms, NATO is quietly building a continent-spanning AI surveillance network, and the Pentagon’s GenAI platform is reaching mass military adoption. The convergence of commercial AI money and live warfighting use is accelerating faster than governance frameworks can keep pace.
Top story: German AI defence startup Helsing raises $1.8bn at an $18bn valuation — Europe’s largest-ever defence tech funding round — signalling a serious challenge to US dominance in military AI.
Helsing Raises $1.8bn, Becomes Europe’s Most Valuable Defence AI Firm
Defense News · Finance
German drone and AI software maker Helsing closed a Series E round of $1.8 billion, valuing the company at $18 billion — Europe’s biggest-ever defence startup funding round — with backing from Goldman Sachs, JPMorgan, and Canada’s CPP Investments. The Munich-based firm builds AI software that fuses data from drones, radar, and satellites into a real-time operational picture for armed forces, and also supplies kamikaze drones to Ukraine. The raise signals that institutional capital is now treating European sovereign defence AI as a category-defining investment, directly rivalling US firms like Anduril.
NATO Leaks Reveal AI ‘Kill Web’ Surveillance Grid Along Russian Border
Yahoo News / Business Insider · Strategy
Internal NATO documents obtained by German outlet BILD reveal a classified initiative codenamed EFDI — the Eastern Flank Deterrence Initiative — to build a unified AI-powered surveillance and targeting network stretching from Finland to Romania. The system would integrate satellites, reconnaissance drones, ground sensors, and robotic platforms, with AI handling real-time data fusion and sharing the resulting picture across all alliance members. NATO describes the strategic logic as ‘deterrence by denial’ — making any Russian attack appear futile before it begins — marking one of the most significant structural shifts in alliance border doctrine since the Cold War.
https://www.yahoo.com/news/world/articles/nato-building-ai-kill-stop-182119271.html
Pentagon’s GenAI.mil Platform Nears 1.7 Million Users, Adds ChatGPT Access
Defense One · Tools
The US Department of Defense’s GenAI.mil platform has recorded nearly 1.7 million users and is set to add OpenAI’s ChatGPT for handling controlled unclassified information from July 2026 onwards. Pentagon officials confirmed the platform’s primary military use-case is aggregating battlefield data for faster decision-making, with the DoD’s chief digital officer noting that human cognition alone cannot keep up with the tempo of modern operations. The scale of adoption — driven partly by the platform’s role during recent combat operations — marks a turning point in how AI tools are normalised across the US armed forces.
Carnegie Endowment: Europe Has a Governance Gap in Autonomous AI Cyber Operations
Carnegie Endowment for International Peace · Regulation
A new Carnegie Endowment paper argues that Europe faces a critical governance gap as AI agents are increasingly used to conduct autonomous cyber operations, with the EU’s AI Act and 2026 Cybersecurity Package both implicitly assuming that developer-side safety alignment is sufficient — an assumption the paper calls dangerously flawed. The report highlights the February 2026 standoff in which the US Secretary of Defense demanded unrestricted military access to Anthropic’s Claude, leading the DoD to designate Anthropic a supply-chain risk when the company refused to remove human-oversight requirements for lethal targeting. For European defence and security practitioners, the paper’s core warning is stark: if US government pressure can erode a major AI provider’s safety commitments, Europe cannot rely on provider-level guardrails as a substitute for its own operational controls.
Palladyne Plans AI Swarming Software on Israeli Combat Drones for US Army
Breaking Defense · Generative AI
US defence AI firm Palladyne is developing new AI swarming software designed to run on battle-proven Israeli drones, with the company’s CEO confirming plans to compete for the US Army’s Long-Range Precision Munition contract. The move represents a direct attempt to bring coordinated autonomous swarm behaviour — where multiple drones collaborate on targeting and attack sequencing without continuous human input — into a major Army procurement programme. The development intensifies scrutiny over human-machine teaming thresholds, arriving just as the Senate considers new rules requiring human oversight of AI-assisted lethal force.