The White House will meet Tuesday with representatives from OpenAI, Anthropic, Google and Meta to discuss a new framework for government review of frontier AI models before public release, according to multiple sources familiar with the invitation. The meeting comes days after both OpenAI and Anthropic disclosed that their AI agents had independently hacked into other companies' systems during routine cybersecurity testing, intensifying calls for tighter oversight.
The framework, outlined in a June executive order signed by President Donald Trump, would give the government access to the most advanced AI models up to 30 days before they are released. Participation is voluntary, but the administration has already taken steps to delay or prevent the release of advanced models for safety reasons.
"The voluntary framework outlined in the June 2nd executive order is complete. Discussions with industry about next steps are underway," a White House official told CNN on Monday.
What the framework covers
The details of the framework will not be released publicly, and many of the standards will be classified, according to the executive order. Major AI companies helped draft the order and framework, but multiple sources familiar with the discussions said there is still uncertainty over final details.
That includes how the administration will define frontier AI models and whether open weight models - those that can be downloaded and customized on a personal computer - will be included. It is also unclear which agency or official will lead the review. National Cyber Director Sean Cairncross, Treasury Secretary Scott Bessent and Commerce Secretary Howard Lutnick have been heading up the initiative, but multiple sources inside the AI companies said the White House has not designated a single person or organization to lead outreach.
OpenAI pointed CNN to a blog post from Chief Global Affairs Officer Chris Lehane that reiterated calls for national AI standards established through Congress. "The Administration's expected action this week on frontier AI could be an important step toward closing the gap between innovation and governance: a clear, credible, national framework for evaluating the most advanced AI systems, with defined criteria, timelines, and a process that allows them to be deployed safely and quickly," Lehane wrote.
Tensions between government and AI companies
The meeting also takes place against a backdrop of strained relations between certain AI companies and federal agencies. The Pentagon designated Anthropic as a "supply chain risk" earlier this year after a disagreement over safety guardrails in the company's models - a designation never before used on an American company. It prohibits the military and its partners from using Anthropic's products.
While Anthropic and the Pentagon are litigating that designation, President Trump told Axios in June that he no longer views Anthropic as a national security threat and said the company has "behaved very responsibly." The White House has been working directly with Anthropic on the executive order and its latest model releases in what sources described as positive and productive discussions.
More than 1,200 staffers at leading AI companies, including Anthropic CEO Dario Amodei, signed an open letter last week urging the U.S. government to create tools that can slow the development of advanced AI models so safety measures can keep pace. "We may have to pace the rate of AI development to give ourselves enough time for society to harden around these new capability levels," OpenAI CEO Sam Altman said on a podcast released last week.
Meanwhile, the European Union's AI Act took effect August 2, giving the European Commission authority to review AI models before release and requiring companies to meet transparency and documentation obligations or face fines of up to 7 percent of annual gross revenue.
Why this matters for government professionals
If you work in the federal government, this framework will likely shape how your agency accesses and deploys advanced AI tools. The review process could affect everything from procurement timelines to which models are available for mission-critical work. The uncertainty over which agency will lead the review, and how open weight models will be treated, means government AI practitioners should expect shifting guidance in the coming months. AI for Government professionals will need to track these developments closely as they determine which systems are approved for official use. Generative AI and LLM models from companies like OpenAI and Anthropic are at the center of this debate, and any changes to their release process will directly affect the tools available to government workers.
Your membership also unlocks:
