White House Weighs Vetting AI Models After Anthropic’s Mythos Revelation

The Trump administration is considering a significant policy shift that would require government oversight for artificial intelligence models before their public release, a move that could reshape the competitive landscape for AI development in the US. This potential reversal from a previously hands-off approach comes as officials grapple with the implications of powerful new models like Anthropic's Mythos, which the company has kept private due to its advanced capabilities in identifying software security vulnerabilities.

According to US officials, the White House is discussing an executive order to create a working group of tech executives and government officials to define review procedures for new AI models. Senior administration officials have already briefed executives from Anthropic, Google, and OpenAI on some of these plans in meetings last week. This marks a stark pivot from the administration's earlier stance, which saw the rollback of Biden-era AI safety evaluation requirements in an effort to foster US competitiveness with China.

The immediate catalyst for this reconsideration appears to be the development of Mythos by Anthropic. The San Francisco-based startup has described the model's potential for a cybersecurity reckoning as a reason for not releasing it publicly. This has raised concerns within the administration about the potential for a devastating AI-enabled cyberattack. The administration is also evaluating the potential for advanced models to be used by the Pentagon and intelligence agencies.

The policy rethink is being driven by White House Chief of Staff Susie Wiles and Treasury Secretary Scott Bessent, who have taken a more active role in shaping AI policy following the departure of Silicon Valley insider David Sacks, a proponent of deregulation. However, their efforts are complicated by an ongoing legal dispute between Anthropic and the Pentagon over a $200 million contract and the terms of AI use by the US military. The market for “Will Anthropic provide Mythos to the US government by April 30, 2026?” is currently priced at 100% YES, but these new regulatory discussions could introduce uncertainty.

The proposed working group would also be tasked with determining which US agencies would have oversight responsibilities. Suggestions have included the National Security Agency (NSA), the White House Office of the National Cyber Director, and the director of national intelligence. Some have even proposed reviving the Biden-era Center for A.I. Standards and Innovation. This move mirrors regulatory processes announced by UK regulators, which assign AI oversight to relevant government bodies.

For investors, this potential shift in US AI policy introduces a new layer of regulatory risk for companies at the forefront of AI development, including Microsoft-backed OpenAI, Google, and Anthropic. While the previous policy of deregulation was seen as a boon for rapid innovation, the introduction of pre-release vetting could slow down development cycles and increase compliance costs. The market has not yet fully priced in these concerns, but any formal announcement from the White House could lead to volatility in tech stocks. The long-term impact will depend on the specifics of the review process and whether it strikes a balance between fostering innovation and mitigating the risks of advanced AI.

This article is for informational purposes only and does not constitute investment advice.