Trump Administration Reverses Course on AI Regulation After Cybersecurity Concerns
Washington, Tuesday, 5 May 2026.
The White House is now considering mandatory government reviews of AI models before public release, marking a dramatic policy shift from Trump’s previous deregulatory stance. This reversal was triggered by Anthropic’s Mythos AI model, which can detect critical software vulnerabilities in programs like Firefox and OpenBSD. The administration fears being blamed for cyberattacks enabled by unrestricted AI technology, prompting discussions for an executive order creating a government-tech industry working group within weeks.
Anthropic’s Mythos Model Sparks Policy Reversal
The Trump administration’s consideration of AI oversight represents a stark departure from its initial approach to artificial intelligence regulation. On January 21, 2025, President Trump rescinded President Biden’s AI executive order, which had required developers of AI systems posing risks to share safety test results with the U.S. government before public release [1][2]. Trump had previously advocated for minimal regulation to help the United States compete with China, dismissing existing rules as “foolish” or “stupid” [3]. However, the emergence of Anthropic’s Mythos AI model in April 2026 has prompted this dramatic policy shift [4][5].
Government Seeks Early Access to AI Capabilities
The White House’s new approach appears designed to address multiple strategic concerns. According to reports, Washington could use stricter oversight to gain early access to powerful new language models, potentially allowing security services or the military to benefit first from these capabilities [1][4]. On April 28, 2026, White House officials briefed executives from Anthropic, Google, and OpenAI about the administration’s considerations for AI oversight [1][2][6]. The discussions include plans for a formal government review process for new AI models, potentially similar to a system being developed in Britain [3].
Public Opinion and Political Calculations
The administration’s policy shift also reflects broader public sentiment regarding AI development. A poll conducted in November 2025 indicated that 50 percent of Republican voters and 51 percent of Democratic voters expressed more concern than excitement about AI’s influence on society [1][2]. This bipartisan anxiety about AI’s rapid advancement provides political cover for the Trump administration’s regulatory pivot, despite the president’s previous promises to eliminate government interference in the AI sector.