OpenAI Seeks Legal Guidance on Coordinating an AI Slowdown

Summary

OpenAI has recently asked members of Congress for guidance on whether AI companies could legally coordinate a slowdown in frontier AI development. The request follows chief scientist Jakub Pachocki’s call for voluntary slowdowns until shared safety standards are established, but legal experts warn that an agreement restricting development could be viewed as limiting output under US antitrust law. Nicholas Felstead, a former AI policy fellow, says the legality would depend on the details, while uncertainty alone could discourage cooperation. A bipartisan, bicameral bill introduced in July would explicitly allow AI laboratories to coordinate on security and safety work; its House version remains with the Judiciary Committee, and supporters say legislation may not advance until after the midterm elections. Some AI leaders argue antitrust concerns are genuine, while others say companies may also be avoiding cooperation because they compete for the frontier-model market, disagree over safety methods, and see AI leadership over China as a national-security priority. Recent warnings about autonomous systems, security incidents, and safeguards failing to keep pace have increased pressure on lawmakers to regulate the industry.