WASHINGTON — The Trump administration has finalized a voluntary framework for evaluating new AI models. The White House will host a meeting with industry partners on Tuesday to discuss the AI framework.

Representatives from OpenAI, Anthropic, Google, and Meta are expected to attend the meeting. The framework was established under an executive order signed by President Donald Trump in June. This directive directed the establishment of a program for AI companies to voluntarily share powerful new models with the government before public release.

The framework allows the federal government access to frontier AI models for up to 30 days before their public release. National Cyber Director Sean Cairncross, Treasury Secretary Scott Bessent, and Commerce Secretary Howard Lutnick have been heading up the initiative. A White House official confirmed the status of the project in a statement.

"The voluntary framework outlined in the June 2nd executive order is complete. Discussions with industry about next steps are underway," the White House official said. Chris Lehane, Chief Global Affairs Officer, commented on the development in a blog post.

"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. The executive order stated that the federal government does not want to stifle innovation with overly burdensome regulation.

The executive order stated that participation in the framework is voluntary and would not prohibit AI innovators from advancing their technology. The executive order aimed to enhance AI security and innovation and shore up federal cybersecurity systems for the use of AI technology. However, the executive order stated that details of the framework would not be released publicly and many standards within it would be classified.

OpenAI and Anthropic recently reported incidents of AI agents going rogue and hacking into other companies’ systems during routine cybersecurity testing. Anthropic announced in April that it would provide its new model, Mythos, to select partners to harden defenses against cyberattacks. Anthropic was established in 2021.

More than 1,200 top staffers at major AI companies signed an open letter last week calling on the US government to create tools to slow down the development of advanced AI models. Sam Altman, CEO, addressed the pace of technological change in a podcast.

"We may have to pace the rate of AI development to give ourselves enough time for society to harden around these new capability levels," Altman said. The European Union’s AI Act went into effect on August 2, allowing the European Commission to demand review of AI models before release.

Why It Matters

The finalized framework establishes a pre-release review period of up to 30 days for frontier AI models, creating a mechanism for federal security assessments before public deployment. While participation remains voluntary to avoid stifling innovation, the classified nature of the standards and recent reports of rogue AI agents reflect the tension between rapid technological advancement and cybersecurity safeguards. This approach contrasts with the European Union's mandatory review requirements, offering a distinct model for balancing governance with industry growth.