President Donald Trump signed an executive order concerning the oversight of artificial intelligence (AI). This action took place shortly after he delayed a White House ceremony due to apprehensions that such a policy might weaken America’s technological leadership. The order provides a framework allowing the federal government to examine national security risks associated with advanced AI systems. This review process is limited to a month before these systems are publicly released. Participating in this vetting is voluntary for AI developers.
The executive order highlights the dual nature of advanced AI capabilities. While they enhance national strength, they also bring up new national security issues requiring cohesive action among government agencies. It remains unclear how this order differs from the previous version Trump chose not to sign on May 21.
Under the new directive, the government has just 30 days to review AI systems, which is a shorter timeframe than some in the industry anticipated. A longer review period might be too demanding for this rapid-moving and competitive field. Trump had previously canceled a planned event with tech executives, expressing concern about retaining the U.S. technological lead, particularly over China and other competitors.
The order forms a voluntary collaboration with U.S.-based tech companies such as Anthropic, OpenAI, and Google. These companies are often referred to as ‘frontier labs’ due to their work on the most advanced AI technologies. The initial cancellation of the signing event disappointed several companies, as executives had planned to attend the May 21 announcement. Eventually, Trump signed the order without public ceremony.
A social media post from the White House explained that the executive order establishes a process for frontier labs to voluntarily share advanced cyber models, aiming to secure critical infrastructure and enhance government cyber defenses. The administration emphasized that it aims to avoid widespread oversight that could stifle innovation or freedom of speech.
Juan Londoño, a policy analyst at the Cato Institute, commented that while the order is not without flaws, it marks progress in preparing for the deployment of advanced AI systems. He commended the voluntary nature of the process but expressed concern over the vague criteria for determining which AI models merit scrutiny. Additionally, he raised worries regarding how ‘trusted partners’ would be chosen for early access, pointing out the significant authority the order grants to the NSA director. Such discretion, Londoño warned, could pave the way for potential government misuse against contentious companies like Anthropic.
Anthropic itself has announced a new AI model named Claude Mythos amid its legal disputes with the Trump administration over contracts with the Pentagon. This development spawned plans for a new AI cybersecurity directive. Treasury Secretary Scott Bessent and outgoing Federal Reserve Chair Jerome Powell flagged concerns to Wall Street executives about Mythos’ ability to detect cybersecurity vulnerabilities. Although access to Mythos has been limited to select partners, Anthropic recently widened this circle to include 150 additional organizations.
Anthropic has welcomed the executive order as a crucial step in bolstering the U.S.’s AI leadership and expressed eagerness to collaborate with the White House on implementing the policy. OpenAI, a competitor, also deemed the policy a forward move. A statement from OpenAI’s chief global affairs officer, Chris Lehane, emphasized the necessity for AI safety frameworks to evolve with input from democratic institutions, technical expertise, and diverse stakeholders to ensure accountability and public confidence.
Democratic Senator Mark Warner, vice chairman of the Senate Intelligence Committee, supported Trump’s policy but criticized the administration for rediscovering the need for such regulations after previously dismantling safeguards established by former President Joe Biden. These actions took place shortly after Trump’s return to the White House.

Garry Kasparov Reflects on AI Advancement
A Race for Experimental Nuclear Reactors: The Push for Innovation vs. Safety Concerns
Florida’s AI Data Center Legislation and Its Implications
California’s New Laws Transform Daily Life
China’s AI Chip Market Sees Shift Amid Nvidia’s Export Challenges
Social Media Safety Advocacy