President Donald Trump privately signed a scaled-back executive order on Tuesday that establishes a voluntary process for tech companies to submit advanced AI models for federal review 30 days prior to public release. The order, which emphasizes non-mandatory compliance, marks a significant reduction from initial proposals calling for a 90-day review period.
National Security Concerns Drive Policy Shift
The decision comes amid growing concerns about cybersecurity vulnerabilities exposed by AI models, particularly following Anthropic’s announcement in April regarding its Claude Mythos system. The startup cited risks associated with the model’s ability to identify software vulnerabilities as a primary reason for limiting its release. Anthropic recently confirmed plans to release Mythos-level models to customers in the near future.
We're leading China. We're leading everybody. And I don't want to do anything that's going to get in the way of that lead.
Trump’s remarks highlight his administration’s focus on maintaining American dominance in AI innovation, particularly in competition with China. The executive order reflects a balancing act between national security interests and the need to avoid stifling domestic technological advancements.
Tech Industry Response
The scaled-back order has been met with cautious optimism by the tech industry, which argued that stringent mandatory reviews could hinder innovation and erode America’s competitive edge. While specific industry responses remain pending, the voluntary framework is seen as a compromise between federal oversight and corporate autonomy.