This week’s defence AI stories are defined by two converging pressures: an accelerating race to deploy autonomous and agentic systems at the operational edge — from AI-directed drone swarms to classified legal workflows — and a mounting transatlantic governance backlash, with both the EU’s landmark cybersecurity action plan and new US Senate legislation forcing hard questions about where human oversight ends. The gap between AI deployment speed and accountability frameworks has never been wider or more consequential.
Top story: The Pentagon quietly awarded a $98.8M AI contract to Casepoint for classified legal and eDiscovery operations, signalling that AI is now embedded across every function of the US military — not just the battlefield.
101st Airborne Uses AI Drone Swarms to Breach Enemy Lines Uncontested
DefenseScoop · Strategy
The 101st Airborne’s 3rd Mobile Brigade Combat Team deployed over 228 one-way attack drones — more than half built in-house for as little as $750 each — to conduct a fully robotic trench breach at Fort Polk, with zero soldiers sent forward until the position was cleared. The exercise marks a doctrinal shift: AI-enabled drone swarms are now being used as the primary breach tool, replacing the most lethal phase of infantry assault. For defence practitioners, the 101st’s model suggests the next procurement wave will centre on cheap, attritable autonomous systems rather than expensive precision platforms.
Pentagon Awards $98.8M AI Contract for Classified Legal Operations
DefenseScoop · Legal
The Pentagon selected Casepoint as the exclusive awardee of a blanket purchase agreement worth up to $98.8 million to deploy AI-powered eDiscovery and legal workflow tools across the DoD’s Office of General Counsel, DISA, and 28 Defence Legal Services Agency offices — all operating at classified Impact Levels 5 and 6. Users reported a 50–75% reduction in time spent on manual document review, with every AI decision logged and auditable. The deal illustrates how AI adoption inside the US military is moving well beyond the battlefield and into sensitive legal, oversight, and FOIA functions — raising novel questions about AI governance in high-stakes, classified environments.
https://defensescoop.com/2026/06/24/pentagon-casepoint-ai-products-enhance-classified-legal-ops/
EU Publishes Landmark AI-Cybersecurity Action Plan With August Enforcement Teeth
European Commission · Regulation
The European Commission published its Action Plan on Cybersecurity and Artificial Intelligence on 7 July, setting nine coordinated actions across three objectives: making frontier AI safe for cyber defence, hardening EU critical infrastructure, and scaling Europe’s sovereign AI capacity. Crucially, from 2 August 2026 the EU AI Act’s enforcement provisions against providers of general-purpose AI models with systemic risk — including cyber misuse — come into force, with fines of up to 3% of worldwide turnover. The plan also commits to a new ENISA-backed European Blueprint giving critical infrastructure operators structured access to frontier AI for cyber defence by Q4 2026, making this the most operationally significant EU cyber-AI policy move to date.