President Donald Trump signed an executive order to oversee artificial intelligence advancements. This decision follows his hesitation to sign a similar policy earlier, fearing it could weaken the U.S.’s technological advantage.
Framework for AI Oversight
The executive order outlines a framework for the federal government to review national security risks associated with the most advanced AI systems. This review process will take place for up to a month before any public release. AI developers can participate voluntarily.
Advanced AI capabilities make our Nation stronger, but also introduce new national security considerations that require coordinated action across executive departments and agencies.
There is some ambiguity about how this order differs from the one Trump declined to sign earlier. It stipulates that the government would have 30 days to review an AI system, which is shorter than industry expectations, allowing for quicker decisions in a fast-paced sector.
Concerns and Industry Response
Trump canceled a previous event with tech executives, citing concerns about the policy’s text. He emphasized the importance of maintaining U.S. leadership in AI.
The directive introduces voluntary collaboration with U.S.-based tech companies such as Anthropic, OpenAI, and Google, known as frontier labs. These companies focus on developing advanced AI systems and were initially planning to attend the signing ceremony.
A White House social media post clarified that the order facilitates frontier labs to voluntarily share advanced cyber models, aiming to secure critical infrastructure and strengthen government cyber defenses.
Analysts and Companies Weigh In
Juan Londoño from the Cato Institute views the order as a step forward for preparing the nation for advanced AI systems, despite its imperfections. He praises the voluntary nature but raises concerns about the director of the National Security Agency having significant discretion in decision-making.
Plans for a new AI cybersecurity directive followed Anthropic’s April announcement of its advanced AI model, Claude Mythos, amid its legal issues with the Trump administration.
Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell held a meeting with Wall Street CEOs, discussing risks posed by the Mythos model’s ability to find cybersecurity vulnerabilities in software.
Anthropic restricts access to Mythos to a small group of trusted partners, including tech companies and banks, and recently expanded this group.
Anthropic, OpenAI, and Google responded positively to Trump’s order, highlighting its role in strengthening AI leadership. Chris Lehane from OpenAI stressed the importance of developing safety frameworks with technical expertise and stakeholder input.
Political Reactions
Democrat Sen. Mark Warner, vice chairman of the Senate Intelligence Committee, appreciated the new policy but pointed out that the administration is reinstating measures dismantled earlier by Trump.
Leave a Reply