Dubai, UAE — US President Donald Trump signed an executive order creating a voluntary framework that allows leading artificial intelligence developers to provide the government access to advanced AI models before their public release, according to France24 and other media reports.
The order permits companies, including OpenAI, Google and Anthropic to voluntarily share their most capable AI systems with the government for cybersecurity testing for up to 30 days before launch, France24 reported.
The measure was prompted by concerns surrounding Anthropic’s Mythos model, which the company has withheld from public release because of its ability to identify vulnerabilities in computer systems, including those used by banks, governments and hospitals, according to the report.
The 30-day review period represents a compromise between an earlier White House proposal that would have required up to 90 days of pre-release access and technology companies that had sought to reduce the period to 14 days, France24 reported.
OpenAI Chief Executive Sam Altman welcomed the order, saying it “gets the balance right.”
“The U.S. should lead on AI by continuing to develop the very best models, making sure they’re safe, and getting cyber tools into the hands of trusted defenders,” Altman said, according to France24.
Kent Walker, Google’s head of public affairs, described the order as “an important step forward” that would help ensure defenders have the AI tools needed to protect national security.
Anthropic also backed the measure, calling it “an important step in strengthening America’s leadership in AI,” according to the report.
The signing followed weeks of debate within the administration. Politico and other media outlets reported that David Sacks, Trump’s AI and cryptocurrency adviser, warned that the proposal could hamper innovation and undermine U.S. competitiveness in the global AI race with China.
Sacks wrote on social media last week that “unnecessary regulation is the biggest threat to innovation in America” and argued that bureaucratic barriers should be removed to support AI development, according to media reports.
The executive order also directs the Treasury Department, the National Security Agency and the Cybersecurity and Infrastructure Security Agency to establish an AI cybersecurity clearinghouse in collaboration with industry and critical infrastructure operators to identify and address software vulnerabilities.
Trump rescinded an AI oversight order signed by former President Joe Biden on his first day back in office. Biden’s 2023 directive required AI companies to share safety-testing results with the government and relied largely on voluntary commitments.
The European Union has taken a more prescriptive approach through its AI Act, which imposes binding requirements on high-risk AI systems, including transparency, safety testing and incident-reporting obligations.
Anthony Aguirre, chief executive of the Future of Life Institute, said the new order was “an important step in the right direction” but argued that voluntary measures alone would be insufficient to manage national security risks posed by advanced AI systems, according to France24.
