White House invites AI leaders from OpenAI, Google, Anthropic and others to review oversight framework

The Trump administration has reportedly finalised a voluntary testing framework designed to evaluate the cybersecurity risks and hacking capabilities of the world’s most advanced artificial intelligence (AI) systems. Citing senior administration officials, a report by The Information (via Axios), following the completion of the policy milestone, the White House has invited technical representatives and leadership from top AI development labs, including OpenAI, Google and Anthropic, to a staff-level meeting on Tuesday to review the proposed oversight procedures.

The initiative follows a June directive from President Donald Trump instructing administration officials to construct standardized benchmarks for assessing high-risk cyber capabilities in frontier models before they are released to the public.