AI capability developments, regulatory frameworks, investment, and the governance contest over artificial intelligence. Published every Friday.
Anthropic's dual-model structure (Fable 5 + Mythos 5) represents a structural shift in how frontier labs manage dual-use capability at release — capability gating is now a product-level feature, not a
Read briefing →Transatlantic regulatory divergence creates forum-shopping opportunities for labs. US labs can avoid EU AI Act obligations by serving EU markets from US infrastructure, while EU labs face dual complia
Read briefing →The US voluntary review mechanism creates a precedent for government visibility into frontier model capabilities without binding enforcement, while the EU Scientific Panel has explicit authority to ad
Read briefing →The EU AI Omnibus political agreement reached 7 May 2026 is the most consequential structural change to the EU AI Act enforcement timeline since the regulation entered force. The 16 to 24 month extens
Read briefing →The EU AI Omnibus political agreement reached 7 May 2026 is the most consequential structural change to the EU AI Act enforcement timeline since the regulation entered force. The 16 to 24 month extens
Read briefing →Anthropic Claude Mythos2 Preview has crossed a threshold for AI-enabled mass zero-day vulnerability discovery, fundamentally altering the offensive-defensive balance in cybersecurity. The scale of zer
Read briefing →Anthropic Claude Mythos2 Preview has crossed a threshold for AI-enabled mass zero-day vulnerability discovery, fundamentally altering the offensive-defensive balance in cybersecurity. The scale of zer
Read briefing →AI models have crossed the capability threshold for competitive vulnerability discovery at scale, with immediate dual-use implications for offensive and defensive cybersecurity operations
Read briefing →The EU AI Omnibus political agreement postponing high-risk AI system compliance to 2027-2028 is a structural admission that the EU standardisation machinery cannot deliver harmonised standards at the
Read briefing →The EU Digital Omnibus package represents a strategic political decision to prioritise industry readiness over regulatory enforcement, creating a 16-24 month window where high-risk AI systems can be d
Read briefing →