OpenAI reportedly asked Congress about a coordinated AI slowdown

WIRED reported, citing people close to OpenAI, that the company asked members of Congress whether an industry-wide slowdown in frontier AI development could violate antitrust law. No public request, named lawmaker or official confirmation of the outreach has been identified.

The legal question hinges on what competing laboratories would coordinate and how. Federal Trade Commission guidance distinguishes exchanges of less competitively sensitive technical information from those involving output or future strategic plans. It warns that an agreement among competitors to limit output may be unlawful in itself. The guidance does not determine whether a specific AI arrangement would be legal.

An archived Justice Department guide distinguished a bare agreement to fix output from a restraint reasonably related and necessary to an efficiency-enhancing collaboration, which could receive a broader rule-of-reason review. The agencies withdrew those 2000 guidelines on December 11, 2024, so they are not a current safe harbor for AI developers.

OpenAI chief scientist Jakub Pachocki has publicly outlined a possible pacing framework, separate from any confirmed agreement. In an OpenAI essay, he paired continued work on alignment and monitoring with coordination to slow future development when needed to build confidence in safeguards. He also advocated broadly mandated safety thresholds enforceable by auditors, government agencies or international bodies.

That proposal is materially different from an operative multi-lab pact. The inspected evidence identifies no participating laboratories, agreed triggers, duration, enforcement mechanism or signed terms. It supports reporting about possible coordination and legal uncertainty, not a claim that frontier labs have agreed to slow development. A separate OpenAI frontier-training pause provides context on a single-lab safety decision, rather than evidence of coordinated action.

Congress is considering a narrower statutory approach. H.R. 9914, introduced on July 23, 2026 and referred to the House Judiciary Committee, would create an affirmative-defense antitrust exemption for good-faith coordination focused exclusively on defined AI security risks. Its protections could cover certain delays or limits, but would require advance written notice to the assistant attorney general and would not cover conduct including price-fixing or market allocation. The bill has not been enacted.

WIRED said OpenAI did not respond to its request for comment before publication.