Under an executive order issued by President Donald Trump, AI companies such as OpenAI and Anthropic are required to submit advanced artificial intelligence models to the government to assess cybersecurity, privacy, “insider risk” and intellectual property protection.
But as the executive order makes clear, AI companies aren’t obligated to do much of anything.
At the end of the section that orders the creation of an evaluation process in which AI companies will send their models 30 days before release, it says: “Nothing in this section should be taken to authorize the creation of a mandatory government license, prior specification, or approval requirement for the development, publication, release, or distribution of new AI models, including boundary models.”
A White House representative did not immediately respond to a request for comment.
Frontier models are advanced AI models that can pose a serious security risk, such as The Anthropic Mytha company that has since been released publicly due to cyber security concerns. US-based AI companies are in a race to train and release artificial intelligence models, especially amid competition from Chinese companies.
But the rapid release cycle of new models has raised concerns that they are not properly tested or controlled before going public, in particular. about their human implications or potential use tools and hacking.
The executive order also includes provisions for agencies, such as the Pentagon and the US Treasury Department, to strengthen their cybersecurity defenses in the next 30 days.
In the next 60 days, agencies, including the Department of Homeland Security and the National Security Agency, are expected to create a framework for testing AI models, although companies are not required to submit them.
(Disclosure: Ziff Davis, CNET’s parent company, in 2025 filed a lawsuit against OpenAI, alleging that it infringed Ziff Davis’ copyrights in training and using its AI programs.)
The ‘look’ of supervision?
Similar executive actions were expected two weeks before the June 2 order, but the final version included several notable changes. According to CNN, the original draft called for a 90-day review period instead of 30 days. However, AI companies involved in the order process, including Anthropic, have reportedly pushed back the long timeline.
CNN also reported that the Commerce Department’s Center for Standards and Technology announcement last month related to AI companies sharing their AI models with the government has disappeared from the agency’s website.
Given the lack of enforcement provisions or mandatory requirements for AI companies, the executive order received a muted response.
“Voluntary frameworks are not enough,” Anthony Aguirre, CEO and president of the AI Safety nonprofit Future of Life Institute, said in an emailed statement to CNET. “We need a mandatory government pre-deployment review process for the most powerful AI systems, allowing the government to prevent the release of systems that pose an unacceptable national security risk.”
The executive order does not specify what happens if the government gets early access to the AI model and a major problem is discovered.
John Thickstun, an assistant professor of computer science at Cornell University, told CNET that this vagueness makes it difficult to understand what the executive order will accomplish.
“Without clear answers to these questions, what I’m learning about this is that it creates the appearance of oversight while largely pursuing a managerial approach to managing AI,” Thickstun said.
