President Donald Trump has signed a new executive order establishing a framework for reviewing the national security implications of advanced artificial intelligence systems before their public release.

The policy introduces a voluntary process that allows the federal government to assess potential risks posed by cutting-edge AI models, with reviews capped at 30 days. The White House said the measure is intended to strengthen national security while preserving America's competitive advantage in the global AI race.

The order relies on cooperation from leading U.S. AI developers, including OpenAI, Anthropic, and Google. According to the administration, the framework focuses specifically on frontier AI systems and critical cybersecurity applications rather than broad oversight of all new models.

Officials emphasized that participation remains voluntary, arguing that mandatory regulation could hinder innovation and technological development. The policy has received support from major technology firms and some lawmakers, who view it as a step toward balancing AI advancement with national security concerns.

However, critics have raised questions about how the government will determine which models qualify for review and the level of discretion granted to national security agencies. The executive order follows growing concerns over increasingly powerful AI systems and their potential ability to identify vulnerabilities in critical infrastructure, financial systems, and cybersecurity networks.