US Weighs Antitrust Guidance on AI Safety, Top DOJ Official Says

The Trump administration is considering whether to update interagency antitrust guidance on cybersecurity to address a new wave of threats posed by artificial intelligence, according to Associate US Attorney General Stanley Woodward.

At an event in New York on Sept. 17, Woodward said that existing guidelines spell out the extent to which companies can coordinate on hacking risks without running afoul of rules prohibiting anticompetitive behavior. While Silicon Valley leaders have called in recent days for an exemption from antitrust scrutiny to allow them to work together on safety, Woodward said that so far no one has contacted his office on the matter.

“To the extent that anybody at any of the frontier labs wants to come see us about a concern, they would get a meeting,” said Woodward, the department’s No. 3 official and the head of its antitrust section. “That meeting hasn’t been requested.”

During the Obama administration, the federal government issued guidance clarifying that companies could coordinate on sharing cybersecurity threat information to better protect networks and deter attacks. A similar approach has been floated by the tech industry for artificial intelligence.

Anxiety over AI’s existential risks has risen since the publication Saturday of an essay by Anthropic PBC Chief Executive Officer Dario Amodei that warned of new threats posed by the technology. Amodei pledged to slow development of cutting-edge models and said the industry would benefit from a narrow antitrust waiver to allow AI companies to work together more on safety issues.

While Woodward signaled some openness to discussing the industry’s concerns, the head of the Federal Trade Commission, which shares jurisdiction over competition enforcement, expressed skepticism about companies’ desire for an exemption.

FTC Chair Andrew Ferguson said Tuesday he would be “deeply suspicious” of requests for relief and said that the AI companies are “asking for barriers to entry” that will insulate their companies from challengers.

Some of the leading US AI developers have already begun safety discussions. OpenAI global policy chief Chris Lehane said Tuesday the company has been working for several weeks on AI risks with Anthropic and Google DeepMind, a unit of Alphabet Inc. He added that their work on security has ample precedent, citing the airline industry as an example. “There are many instances over time where companies are trying to help each other on safety,” he said.

During his appearance on Thursday, Woodward said the administration was watching the issue “very closely” but was waiting still to hear from the industry.

As of now, he said, officials “have not had any outreach from companies.”