White House finalises but withholds its voluntary frontier-AI model review framework
President Trump's 2 June executive order (EO 14409) gave agencies 60 days — to 1 August — to build a voluntary process letting AI labs submit frontier models for government vetting before release. The deadline passed with nothing published; the White House then confirmed on 3 August the framework was in fact finished, but said it will not release the document or disclose which companies have seen it, reviewing the draft behind closed doors on 4 August with Meta, Nvidia, Microsoft, OpenAI and Anthropic among others. Because the order explicitly bars the framework from operating as a mandatory licensing or preclearance requirement, participation is voluntary and enforcement leverage is unclear. Worth flagging to any US-facing client as a live contrast with the EU's own Code of Practice below — also voluntary, but published in full with a running signatory list: transparency of the rulebook itself, not just compliance with it, is now a live governance-design question on both sides of the Atlantic.
Source: Fortune
Commission confirms ~190 signatories to the AI-Generated Content Transparency Code ahead of Article 50
The Commission confirmed on 31 July that roughly 190 organisations — including OpenAI, Anthropic, Google, Microsoft, Meta, Mistral and Cohere on the provider side, and deployers such as Getty Images, Lufthansa and Iberdrola — have signed the Code of Practice on Transparency of AI-Generated Content, the voluntary instrument operationalising Article 50's marking and labelling duties. About half the signatories are small or recently founded firms; the Commission's own framing is that signing gives a 'streamlined, predictable and legally certain pathway' to demonstrate compliance rather than having to prove it from scratch to a regulator. Two industry task forces launch in September 2026 to shape implementation, so signing now also buys a seat at that table. The list stays open and updates on a rolling basis (last updated 5 August) — worth checking again before advising a client the Code isn't relevant to their AI stack.
Source: European Commission — Shaping Europe's Digital Future
Unsure which of these developments applies to your AI systems?
Begin in writing →This briefing is general information, not legal advice, and does not create an advisor–client relationship. Summaries are original; follow source links for the full record. Adesanya AI Advisory — Abdulwahab B. Adesanya, Barrister-at-Law (Nigeria).