OpenAI has spent recent weeks asking members of Congress for clarity on whether a coordinated industry-wide slowdown in frontier AI development would be lawful, according to people familiar with the company's discussions. The question matters to OpenAI because substantive safety coordination among AI labs could run into antitrust law, those people say. That legal risk, they add, is a major hurdle to getting large technology companies to join such an effort.
OpenAI chief scientist Jakub Pachocki made the company's position public last weekend in a blog post. He wrote that coordinating to slow future development is part of the best path forward for AI research, calling it key to keeping self-improving AI systems safe. In the near term, he expects voluntary slowdowns to become common until shared safety standards are in place.
Not everyone agrees such coordination is permissible. Nicholas Felstead, an assistant director at the Australian Competition and Consumer Commission and a former AI policy fellow at the Center for Law & AI Risk, wrote in a March article that a coordinated pause in AI development could amount to companies limiting output and potentially violate the Sherman Antitrust Act. Felstead wrote that the outcome would depend entirely on the specifics of any agreement, and that even if most safety collaborations survived antitrust review, legal uncertainty could deter companies on its own.
Congress has shown early interest in the issue. In July, a bipartisan, bicameral group of lawmakers introduced the Collaboration on Adversarial Threats and Security Risks Act, which would expressly allow AI labs to coordinate on security and safety work without antitrust exposure. The House version went to the Judiciary Committee but has not been taken up. Caleb Knapp, director of government affairs at the nonprofit AI Policy Network, which backs the bill, said it would create legal channels for labs to jointly address safety and security incidents. He said Congress has a growing appetite to act on AI safety, though legislation may wait until after the midterm elections.
Not all AI leaders accept the antitrust concern as the real obstacle. John Schulman, an OpenAI cofounder who is now chief scientist at the rival lab Thinking Machines, wrote on X this week that OpenAI and Anthropic should stop feuding and develop a pacing proposal together. He said antitrust prohibits certain agreements but does not bar jointly developing a proposal, calling the legal argument fake.
The debate follows a summer of rising public alarm about the AI race. This week, former Anthropic and OpenAI researcher Jacob Coxon publicly warned that AI developers were putting humanity at risk. Recent security incidents, including OpenAI's agents hacking Hugging Face, have shown safeguards failing to keep pace with model capabilities, prompting urgent calls from many lawmakers for regulation.
More AI news from TechManNews.
