WASHINGTON, May 4 (Reuters) — The White House is weighing whether to introduce formal government reviews of newly developed artificial intelligence models, the New York Times reported on Monday, citing officials familiar with internal deliberations.
According to the report, federal policymakers are considering an executive order to establish an AI working group that would bring together technology executives, cybersecurity experts and government officials to evaluate oversight procedures for advanced models prior to public release.
A White House official declined to confirm or deny the story, saying, "Any policy announcement will come directly from the president" and that discussion of potential executive orders remains "speculation."
Why the Shift?
The Times linked the possible policy shift to concerns about Anthropic's recently unveiled model, Mythos. Cybersecurity experts have warned that Mythos's advanced coding and vulnerability-detection capabilities could accelerate or enable more complex cyberattacks by identifying weaknesses and suggesting exploit strategies.
Experts told the paper: Mythos's higher-level coding abilities may give it an unprecedented capacity to locate cybersecurity vulnerabilities and outline ways to exploit them.
Political Context
The move would represent a notable change for President Donald Trump, who has largely promoted a hands-off approach to AI. In July, his administration released an AI blueprint that sought to loosen certain regulatory constraints and expand AI exports to allies to help maintain U.S. technological advantage over China.
On his first day in office in 2025, President Trump revoked a 2023 executive order issued by President Joe Biden, which had aimed to reduce the risks AI poses to consumers, workers and national security. Biden's order required developers of AI systems that could threaten national security, the economy, public health or safety to share safety-test results with the federal government—using authorities similar to the Defense Production Act—before public release.
Separately, the White House in March unveiled policy guidance for Congress urging lawmakers to enact federal legislation to pre-empt a patchwork of state rules, protect children and vulnerable communities, and address energy impacts related to rapidly expanding AI systems.
It remains unclear whether any executive order will be issued and what its final scope would include. Administration officials, technology companies and civil-society groups are likely to continue discussions as policymakers weigh how to balance innovation, competitiveness and public safety.
Reporting by David Shepardson and Alex Alper in Washington and Chandni Shah in Bengaluru; Editing by Sonali Paul.
