Trump Eases AI Oversight: A Controversial Executive Order

A Relaxed Executive Order for AI

President Donald Trump recently signed an executive order that modifies the review rules for artificial intelligence (AI) models in the United States. This order allows tech companies to voluntarily submit their AI models for federal government review 30 days before their official release.

This decision comes amid growing concerns regarding the cybersecurity capabilities of certain AI models, notably Claude Mythos, developed by the company Anthropic. This model has raised alarms among industry experts due to its potential effectiveness in detecting software vulnerabilities.

Discussions and Adjustments

The executive order signed by Trump follows repeated discussions between the federal government and leading AI companies, including Anthropic. In April, this startup had already limited the release of Claude Mythos, fearing that its advanced capabilities could pose cybersecurity risks.

Anthropic, which recently filed for a confidential IPO, announced last week that it plans to soon release Mythos-level models for its clients. However, a spokesperson for the company did not immediately respond to requests for comment regarding the executive order.

A Voluntary and Shortened Review

The executive order reduces the initially planned review period from 90 days to just 30 days. This review is not mandatory, but it offers companies the opportunity to voluntarily submit their advanced AI models for early evaluation.

Trump explained to reporters that he had delayed signing the order out of concern for compromising the United States' lead over China in the field of AI. "We are ahead of China. We are ahead of everyone," he stated on May 21, emphasizing his intention not to jeopardize this technological advantage.