Trump Administration Finalizes AI Review Framework, Sets August 1 Deadline

According to The Information, on July 28, the Trump administration is finalizing a voluntary framework requiring AI companies to submit their most advanced models for government review before public release. OpenAI, Anthropic, and Google received the draft framework from the White House's National Cyber Director's office approximately two weeks ago and jointly submitted revisions.

The framework, mandated by an AI-related executive order Trump signed in early June, aims to establish standardized review processes for powerful new AI systems capable of identifying network vulnerabilities. As the August 1 deadline approaches, some AI companies have expressed concerns in White House meetings about how the framework will define frontier models and whether it will distinguish between closed-source and open-source systems. The administration insists compliance remains voluntary.