This week’s Defence & Security AI landscape is defined by rapid operationalisation: the US military is embedding agentic AI directly into classified networks and core IT infrastructure, while Europe accelerates its own AI-defence industrial base under the €800 billion ReArm Europe umbrella. Forrester’s new threat intelligence report captures the other side of the equation — nation-state actors are now deploying autonomous AI to attack at speeds that outrun human defenders, making the arms race both institutional and technical.
Top story: The US Army has issued a landmark RFI seeking agentic AI platforms deployable on its most sensitive classified networks — JWICS and SIPRNet — signalling a decisive shift from AI experimentation to front-line intelligence operations.
US Army Seeks Agentic AI for Top-Secret Intelligence Networks
Intelligence Community News · Strategy
The US Army, in cooperation with DEVCOM Army Research Laboratory, has issued a formal market survey seeking agentic AI platforms that can be deployed on classified networks including JWICS and SIPRNet at Impact Level 6. The platforms must enable intelligence analysts to build, deploy, and govern AI agents capable of reasoning and taking actions inside accredited classified environments — including on edge and disconnected battlefield systems. This marks a significant escalation from AI pilots to operational procurement on the military’s most sensitive infrastructure.
https://intelligencecommunitynews.com/agentic-ai-platforms-for-army-intelligence-rfi-released/
Empower AI Wins $255M Pentagon IT Contract with AI Automation Push
GovCon Wire · Finance
Empower AI has secured a potential five-year, $255.4 million recompete contract from the Defense Information Systems Agency to provide enterprise IT support for the Pentagon and the National Capital Region, serving over 300,000 defence personnel and civilian analysts. The award signals a deliberate move beyond basic service delivery: the contract explicitly tasks Empower AI with introducing AI-driven automation to reduce manual effort, accelerate incident resolution, and cut operational costs across one of the most complex IT environments in the federal government. It underscores how AI is now embedded in the foundational infrastructure underpinning US defence operations.
https://www.govconwire.com/articles/empower-ai-255m-disa-j6-service-delivery-recompete-contract
Forrester: Nation-States Deploy Agentic AI at Speeds Outrunning Human Defenders
Cybersecurity Insiders · Risk
Forrester’s newly released Top Cybersecurity Threats in 2026 report identifies near-autonomous nation-state attacks as the foremost CISO challenge of the year, warning that adversaries now deploy agentic AI to automate and scale exploitation faster than human security teams can respond. The report maps five converging risk categories — including rogue AI agents, AI identity sprawl, and software supply chain exposure — that have materially expanded the threat surface beyond what traditional perimeter-based threat intelligence covered even two years ago. For defence and security practitioners, the report is a direct call to extend governance frameworks to AI agents and automated identities as primary attack vectors.
https://www.cybersecurity-insiders.com/forrester-top-cybersecurity-threats-2026/
EU’s ReArm Europe Funds 57 AI and Drone Projects Worth €1 Billion
Bruegel · Strategy
Europe’s ReArm Europe programme is beginning to deliver at scale: the European Defence Fund’s 2025 call — announced in April 2026 — produced 57 collaborative R&D projects totalling €1.07 billion, spanning AI, cyber, drones, counter-drone systems, and air and missile defence. Underpinned by €150 billion in SAFE loans and relaxed fiscal rules for 17 EU member states, the programme represents the most significant reorientation of European defence industrial policy in decades. For UK and European defence technology firms, this marks a structural opening: France’s defence procurement agency alone has already awarded a €240 million contract for AI processing of military data and ordered 1,000 tactical drones from Paris-based startup Harmattan AI.
Darktrace: Enterprise AI Expansion Has Opened a Vast New Attack Surface
Darktrace · Risk
UK-headquartered Darktrace has released its State of AI Cybersecurity 2026 report, based on a survey of 875 IT and security professionals across the US, UK, Australia, and Germany conducted in late May and early June 2026. The report finds that as enterprise AI has reached every facet of organisational structure in 2026, it has simultaneously created a new and rapidly expanding attack surface that adversaries are actively exploiting with their own AI tools. The findings are directly relevant to defence and security operators managing hybrid civilian-military IT environments, where the speed of the AI-versus-AI threat cycle is outpacing traditional security operations centre workflows.
https://www.darktrace.com/blog/the-state-of-ai-cybersecurity-2026