Trump signs executive order to control AIs before they are released

President Donald Trump signed an executive order Tuesday created a “voluntary system” for AI companies to share their models with the government before they are released to “enhance safe technology and strengthen cybersecurity.”

The order says the US AI industry has succeeded in part “because we refuse to restrict this technology with overly burdensome regulations,” but it also recognizes that new AI capabilities come with security risks. Therefore, it is ordering several government agencies to come up with a plan to “evaluate the advanced cyber capabilities of AI models” before they are released to the public. Companies may have the discretion to share their samples with the government before they are released, but they may gain access to some confidential information if they choose to do so. It is also necessary for the government to plan for AI cyber security, especially for critical infrastructure.

This law comes after Trump put off at the last minute to sign a pre-arranged order that he complained they can “step into the path” of competing with China. Even the previous version would have allowed AI companies to voluntarily share their models 14 to 90 days before release, according to The New York Timesthe current model asks companies to share their samples up to 30 days before releasing them to the public. Google, Microsoft, and xAI agreed last month allowing the review to be released by the Commerce Department’s Center for AI Standards and Innovation (CAISI).

OpenAI and Anthropic they had already agreed to share their samples with CISI back in 2024 under President Joe Biden as part of the Biden campaign. AI security protection. But until recently, the Trump administration downplayed security concerns and took action under former White House AI chief David Sacks. The order signed Tuesday makes it clear that it should not be taken as an official endorsement or guarantee. However, it shows the willingness of the Trump administration to use it to regulate the AI ​​industry.

One of those changes may be Anthropic’s limited April release of its powerful Mythos version, which the company said revealed “multiple vulnerabilities, including some on both major and non-open systems.” Legends were also seen make room for conflict resolution between Anthropic and Administration, following his legal battle with the Pentagon on the use of AI in autonomous weapons and mass surveillance.

The newly signed law has so far been praised even by groups that have advocated for government AI restrictions. “The White House is legally obligated,” America for Responsible Innovation President Brad Carson said in a statement, adding that the plan shows the Trump administration is taking the threat of AI seriously. The CEO of the Alliance for Secure AI, Brendan Steinhauser, said his group is “pleased to see that the Trump administration is taking these types of threats seriously,” and both Steinhauser and Carson urged Congress to enact legal protections.