International policy desk
International
The EU AI Act, UK regulators, and cross-border AI governance — Brussels, London, Geneva.
International
The Commission's July 7 plan tasks ENISA with a structured-access blueprint for frontier models and stands up an EU testing platform by year-end — leaning on existing statutes.
By Tatiana Petrova, International policy · Jul 9, 2026 · International
Why it matters: Brussels is betting the AI Act, NIS2, CRA, DORA and the Cyber Solidarity Act are enough — and that the real gap is operational access to US frontier models, not statutory authority.
Read the brief → International
The General Court's July 8 dismissal of Apple's DMA challenges creates a binding sequencing rule barring all six gatekeepers from pre-emptive litigation — leaving Google 18 days and no legal escape.
By Tatiana Petrova, International policy · Jul 9, 2026 · International
Why it matters: Google can no longer front-run the July 27 Android AI and Search data-sharing decisions with abstract court challenges — the Commission's specification orders will land, and Gemini's system-level advantage on Android is on the clock.
Read the brief → International
The Council gave final approval to the Digital Omnibus on June 29. Publication in the Official Journal is imminent — freezing a two-track enforcement calendar.
By Tatiana Petrova, International policy · Jul 6, 2026 · International
Why it matters: Brussels blinked on high-risk AI compliance but held the line on GPAI enforcement and content-transparency rules — giving industry an 18-month reprieve without gutting the Act.
Read the brief → International
The Council of the EU gave final approval to the AI Act simplification package on June 29, deferring high-risk rules to December 2027 while transparency obligations still hit August 2, 2026.
By Tatiana Petrova, International policy · Jul 3, 2026 · International
Why it matters: Small and mid-size businesses just got a 16-month reprieve on the hardest rules and an expanded compliance-lite framework — but the transparency clock still runs out this August.
Read the brief → International
Council gave final green light June 29 to the AI Act simplification package, delaying high-risk obligations to December 2027 — but transparency rules for chatbots and AI-generated content remain on for August 2, 2026.
By Tatiana Petrova, International policy · Jul 2, 2026 · International
Why it matters: Brussels bought industry 16 more months on high-risk AI compliance, but chatbot disclosure and AI-content labeling duties take effect in 31 days — the deadline nobody moved.
Read the brief → International
The EU's chatbot-disclosure and AI-content-labeling rules apply in 38 days, with final guidelines still pending and the AI Omnibus leaving the date untouched.
By Tatiana Petrova, International policy · Jun 25, 2026 · International
Why it matters: Every company whose AI outputs reach EU users — wherever it sits — faces chatbot disclosure and content-marking obligations on August 2, with fines up to €15 million, and the Commission's final implementing guidance still isn't out.
Read the brief → International
The European Commission released its voluntary Code of Practice on marking and labelling AI-generated content on June 10, giving providers and deployers a Commission-backed path to comply with Article 50 of the AI Act.
By Tatiana Petrova, International policy · Jun 13, 2026 · International
Why it matters: Signing the voluntary code is now the cleanest route to demonstrating compliance with binding Article 50 transparency rules that hit on 2 August 2026 — and non-signatories should expect more scrutiny from national market surveillance authorities.
Read the brief → International
Council and Parliament reach political agreement on amendments. New prohibitions on nudifier apps. Regulatory sandbox deadline postponed to August 2, 2027.
By Tatiana Petrova, International policy · May 7, 2026 · International
Why it matters: The omnibus is Europe's first major adjustment to the AI Act since GPAI obligations took effect last August. The transparency-grace-period cut from six months to three accelerates the labelling requirement on synthetic content.
Read the brief → 
The AI policy news desk