Associate Attorney General Stanley Woodward clarified the Justice Department's stance on antitrust questions surrounding artificial intelligence development in remarks at Fordham University in New York on September 17.
Cooperation among AI companies on cybersecurity issues does not appear to violate antitrust laws, Woodward said at the 53rd annual Conference on International Antitrust Law and Policy. The department is actively reviewing whether to update its existing guidance on the subject and remains open to meeting with companies, though leading frontier AI labs have not yet requested such discussions.
Woodward's comments provided more specific direction after Attorney General Todd Blanche declined to address the issue two days earlier. On Tuesday, Blanche would not confirm whether companies needed an antitrust waiver to coordinate on AI safety, saying only: "I can't address that. What I can say is that to the extent that those companies are looking to regulate themselves or fix problems that they have, they should do that, but working with us is unnecessary."
The pressure for clarity intensified after Anthropic researcher Jacob Coxon resigned last week, citing concerns that AI systems could pose existential risks. Anthropic CEO Dario Amodei has publicly called for antitrust permission to coordinate a slower pace of AI advancement, arguing that safety measures require time to catch up with increasingly powerful systems.
Amodei's proposal highlighted a central tension: how can companies collaborate on shared risks while adhering to laws designed to maintain competition. Existing federal guidance, established in a 2014 joint policy statement by the Justice Department and the Federal Trade Commission, distinguishes between sharing cybersecurity information and exchanging sensitive business data. Legitimate exchanges about cyberthreats differ from sharing information about prices, output, or business plans, which could raise antitrust concerns.
President Donald Trump has encouraged the AI industry to maintain rapid development rather than reduce its pace, adding a political dimension to the coordination debate.
Woodward's distinction between cybersecurity and broader development coordination may allow firms to address specific security vulnerabilities together without immediate antitrust scrutiny. Any agreement among rivals to limit the pace of AI development could still face significant antitrust challenges.
