President Trump signed an executive order Tuesday that asks AI companies to let the government review major new artificial intelligence models 30 days before releasing them to the public. It applies to models the government will classify as "frontier models" — in other words, large-scale, general-purpose AI models like OpenAI's GPT and Anthropic's Claude.
"Advanced AI capabilities make our nation stronger, but also introduce new national security considerations that require coordinated action across executive departments and agencies," the order stated. "My Administration will continue to work closely with industry to ensure that the best and most secure technology is deployed rapidly to confront any and all threats to our country."
The executive order comes about a month and a half after several big-bank
The order further directs several government agencies to help community banks, as well as state and local authorities, rural hospitals and local utilities, gain access to cybersecurity tools and services including frontier models.
This was welcome news to community bankers, who worried about their ability to protect themselves against the security risks of Mythos while large banks got an early look at it.
"ICBA and the nation's community bankers thank the administration for expressly recognizing community banks as a critical infrastructure component and adopting elements of our
The president's inclusion of community banks "reflects the priorities we have advanced in meetings with the administration and in our action plan, particularly our calls for adopting effective AI-driven cybersecurity, modernizing security operations, building an AI-ready workforce, and establishing a voluntary clearinghouse in collaboration with the AI industry and critical infrastructure operators to facilities AI cyber readiness and response," she said.
The order instructs the Secretary of the Treasury, in consultation with the National Cyber Director, the Secretary of War, and the Secretary of Homeland Security to form an AI cybersecurity clearinghouse, in voluntary collaboration with the AI industry and operators of critical infrastructure, that "coordinates and deconflicts scanning for software vulnerabilities, discovers and validates such vulnerabilities, and coordinates and prioritizes remediation and distribution of vulnerability patches."
The order also directs the Attorney General to prioritize the enforcement of federal criminal laws against anyone who uses AI to illegally access or damage a computer or commit a data breach.
The International Center for Law & Economics gave the executive order high marks, yet said guardrails need to be established around the new government process.
The order "takes a more sensible approach to AI security by rejecting mandatory licensing, keeping pre-release engagement voluntary and emphasizing defensive deployment," the Center's Director of Innovation Policy Kristian Stout said in a statement.
But Stout warned that its voluntary review process could still become a de facto gatekeeping regime that favors incumbents, slows innovation or gives future administrations too much leverage over model developers.
"The order gets important things right," Stout said. "It treats AI as a national-security issue without pretending every model should need a federal permission slip. That is a meaningful improvement over proposals that would turn frontier AI into a licensed industry."
Stout worries that the new government framework could be used to "pressure developers on lawful content, pull them into surveillance beyond statutory limits, or slow the release of defensive tools Americans will need."
On the other hand, some tech industry participants said the order does not go far enough.
"Voluntary compliance is nothing more than theater," said Mark Montgomery, founder and CEO of AI company Kyield, in a LinkedIn post. "Obviously intended to placate the increasingly hostile critics of the lack of any relevant governance over LLMs whatsoever, including a growing percentage of both major political parties."
Brendan Steinhauser, CEO of the nonprofit Alliance for Secure AI, said he was pleased the Trump administration is taking the risks of models like Mythos seriously.
"However, we know that Big Tech will still try to cut corners on safety and security," he said. "The next AI models will be even more powerful and will pose even bigger threats to our country than Mythos. These companies need oversight and cannot be trusted to do the right thing voluntarily." He called for Congress to codify the White House's EO with legislative action and create a legal framework that makes federal government review of advanced AI models mandatory.
