This week’s defence AI landscape is defined by a surge in autonomous systems investment and policy acceleration on both sides of the Atlantic. The US White House issued a landmark executive order hardening AI-enabled cyber defences, while a $200M funding round for autonomous counter-drone firm Allen Control Systems underscores the commercial momentum behind AI weaponry. Governments and investors are racing ahead of international governance frameworks, leaving the ethics of human-out-of-the-loop lethal force as the defining unresolved tension.
Top story: Trump’s new AI security executive order puts frontier AI models under voluntary pre-release government review and orders a 30-day cyber defence overhaul across all national security systems.
Trump Signs AI Security Order Covering All National Security Systems
White House / Perkins Coie · Regulation
On 2 June 2026, President Trump signed Executive Order 14409, directing federal agencies to upgrade cyber defences against AI-enabled threats within 30 days and establishing a voluntary framework requiring AI developers to give the government up to 30 days’ early access to frontier models before public release. The order explicitly covers National Security Systems and Department of War information systems, and directs CISA to expand AI-enabled defensive tools for critical infrastructure operators including utilities and community banks. For defence practitioners, the aggressive timelines mean concrete classified benchmarking criteria and procurement implications could emerge as early as July 2026.
Autonomous Counter-Drone Firm Raises $200M at $2.2Bn Valuation
The Defense Post · Finance
Allen Control Systems (ACS) closed a $200 million Series B round, valuing the Austin-based firm at $2.2 billion, to scale its Bullfrog autonomous weapon station — an AI, computer vision, and robotics platform that bolts onto existing weapons like the M240 machine gun to defeat drone threats. Bullfrog is already deployed with the US Army and Navy and achieved a reported 100% success rate at the Pentagon’s T-REX 26-1 counter-drone testing event. The raise signals that investors and the military are betting heavily on AI-driven, cost-effective alternatives to traditional missile-based air defence as drone threats proliferate globally.
https://thedefensepost.com/2026/06/09/allen-control-systems-funding/
DoD’s New CMMC-for-AI Framework Forces Contractors Into Security Compliance
Crowell & Moring LLP · Risk
The US FY2026 National Defense Authorization Act directs the Pentagon to build a dedicated cybersecurity framework for all AI and machine learning systems it acquires, covering risks such as model tampering, adversarial prompt injection, and jailbreaks — and to embed it into the existing CMMC contractor certification programme. Any company developing, deploying, storing, or hosting AI for the DoD will be classified as a ‘covered entity’ subject to the framework once finalised, with a Congressional status update due by 16 June 2026. Defence supply-chain professionals should treat this as an emerging compliance obligation on par with the years-long CMMC rollout that caught many contractors unprepared.
EU AI Act’s Military Exemption Creates Lethal Weapons Governance Gap
European Scientific Journal · Regulation
A peer-reviewed analysis published in the European Scientific Journal highlights that the EU AI Act explicitly excludes military and defence applications, potentially creating a significant regulatory gap for lethal autonomous weapon systems (LAWS) operating in European theatres. The study argues that the transition from automation to full algorithmic autonomy challenges core International Humanitarian Law principles — particularly the requirements of distinction and proportionality — without any EU-level ruleset to enforce them. For UK and European defence professionals, this gap is particularly acute given ongoing debates about autonomous targeting and the absence of a binding international treaty on killer robots.
US Army Awards AI-Drone Threat Detection Kit in First-of-Kind Contract
SEC Filing / Safe Pro Group · Strategy
Safe Pro Group has been awarded a US Army subcontract to deliver a turnkey battlefield intelligence ‘kit’ pairing its edge-compute AI Navigation Observation & Detection Engine (NODE) with Red Cat Black Widow drones, marking the first such integrated kit order of its kind. Powered by Safe Pro’s patented SPOTD technology — trained on one of the world’s largest real-world drone-based imagery datasets — the system can instantly detect small threats including landmines, cluster munitions, and ambush drones. The contract is a concrete example of agentic, edge-deployed AI moving from experimentation to field delivery within the US Army’s rapidly expanding autonomous systems procurement pipeline.
https://www.sec.gov/Archives/edgar/data/0002011208/000149315226026778/ex99-1.htm