Trump plan to combat AI risks hinges on Big Tech pals policing themselves

Two dozen technology firms voluntarily committed Tuesday to implementing controls recommended by the White House, including independent safety audits testing whether companies' internal controls, monitoring and detection work, according to an agreement reported amid escalating AI security incidents that caused OpenAI to halt training and pause releases.
External reviews were to focus on risks related to cybersecurity, biosecurity, chemical threats and unintended actions by AI models. Firms also agreed to meet regularly to discuss best practices and set common AI safety standards and benchmarks. Signers included Anthropic's Dario Amodei, OpenAI's Sam Altman, SpaceXAI's Elon Musk, Nvidia's Jensen Huang, Meta's Mark Zuckerberg and Alphabet/Google's Sundar Pichai.
Nothing new is legally required of firms, and some, including Anthropic, Google and OpenAI, had already committed to external audits, The Information reported. Many firms have also voluntarily committed to secretive government safety testing. Trump insisted the deal is morally binding despite carrying no legal weight, Bloomberg reported, telling reporters it is almost like a constitution in a way, adding that the biggest people in the world signed it and he signed it as president.
The agreement said it may eventually make sense to codify the recommendations. Few details have been released on how the audits would work; Trump posted a summary on Truth Social. Bloomberg reported Trump appears to be politicizing the selection process, and some firms are uncertain what might qualify as a third party, though some prospective reviewers have met the Center for AI Standards and Innovation.
AI industry sources told Axios they were concerned self-policing would not solve the safety problem. OpenAI's Safety and Security Committee chair Zico Kolter told NBC News the company is adapting and needs governance structures that rise to the needs, saying recent incidents showed capable AI agents can act in ways developers did not intend.
States are also involved: OpenAI made commitments to California and Delaware attorneys general during restructuring, and both states are investigating whether governance failures contributed to loss-of-control incidents. The FTC announced a broad investigation into potential consumer harms from rogue AI agents, The New York Times reported. Sen. Bernie Sanders accused Trump of favoring industry interests.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
Trump gets dozens of AI firms to agree to voluntary safety tests.