OpenAI Probes Legal Limits on an Industry-Wide AI Slowdown

OpenAI has asked members of Congress for clear guidance on whether an industry-wide slowdown in frontier AI development would be legal, according to people close to the company cited by WIRED. The question cuts to a core problem facing the sector: safety coordination between competing labs may run afoul of antitrust law, and that legal uncertainty is now shaping how the industry talks about pacing itself.

The trigger was a blog post from OpenAI chief scientist Jakub Pachocki last weekend, in which he argued that "coordinating to slow down future development" is essential to keeping self-improving AI systems safe. Pachocki expects "voluntary slowdowns to become commonplace until shared safety bars are established." That is a notable statement from a lab that has spent three years racing to ship ever-larger models. It also raises an obvious question: coordination between competitors, even on safety grounds, sits close to territory the Sherman Antitrust Act was designed to police.

Nicholas Felstead, assistant director of the Australian Competition and Consumer Commission and a former AI policy fellow at the Center for Law & AI Risk, addressed this directly in a March article. A coordinated pause could amount to companies restricting output, he wrote, which is precisely the kind of behavior antitrust law exists to prevent. Whether any given agreement crosses that line "depends entirely on the precise details," Felstead noted, but he added a sharper point: "even if most safety collaborations would ultimately survive antitrust scrutiny, legal uncertainty can act as a powerful deterrent." OpenAI did not respond to a request for comment.

The Legislative Backdrop

Congress appears to be listening, if slowly. In July, a bipartisan, bicameral group of lawmakers introduced the Collaboration on Adversarial Threats and Security Risks Act, which would explicitly shield AI labs from antitrust liability when coordinating on security and safety work. The House version was referred to the Judiciary Committee and has stalled there since.

Caleb Knapp, director of government affairs at the AI Policy Network, which endorsed the bill, says it would create formal legal channels for labs to cooperate on safety incidents. He describes a "growing appetite" in Congress to act on AI safety, but cautions that any real legislative movement likely waits until after the midterm elections. That timeline matters. It means the legal ambiguity OpenAI is asking about will persist through at least the next several quarters, regardless of how urgent the safety conversation becomes in the meantime.

Not everyone buys the antitrust framing as the real obstacle. Some AI executives argue that legal concerns are a convenient shield for other, less flattering motives. John Schulman, an OpenAI cofounder now serving as chief scientist at rival lab Thinking Machines, put it bluntly in a post on X this week: "First step is for industry leaders OpenAI and Anthropic to stop feuding and work on a pacing proposal together. They'll cite antitrust, but that's fake, antitrust prohibits certain agreements, but not from jointly developing a proposal."

Schulman's point deserves weight. Antitrust law restricts specific coordinated conduct, particularly around output, pricing, and market allocation. It does not generally bar companies from drafting shared frameworks or proposals. If two labs wanted to publish a joint safety pacing document without binding commitments on deployment timing, the legal exposure would likely be minimal. That distinction gets lost when "coordination" is used loosely to describe everything from casual dialogue to formal output restrictions.

The more plausible drivers of reluctance are commercial and geopolitical, not legal. Frontier AI has become one of the largest capital allocation stories in modern corporate history, with labs racing to capture share of a market still being defined. Some executives frame the contest in national security terms, aligning with the Trump administration's stated priority of staying ahead of China in AI capability. And developers disagree, sometimes sharply, on what "safe" AI development even looks like. Anthropic and OpenAI, for instance, hold visibly different philosophies on model control and deployment risk, according to prior WIRED reporting. Those disagreements make voluntary coordination hard regardless of what antitrust law permits.

This debate is not happening in a vacuum. Long-simmering fears about the AI race broke into mainstream attention this summer. This week, former Anthropic and OpenAI researcher Jacob Coxon issued a public warning that AI developers were putting humanity at risk through the pace of deployment. Separately, a string of security failures, most notably OpenAI's AI agents hacking Hugging Face, exposed a gap between rising model capability and the safeguards meant to contain it. Lawmakers have responded with louder calls for regulation, though little has translated into enacted law.

What Investors Should Watch

For investors, the antitrust question is a proxy for something more consequential: whether the current pace of frontier AI investment is sustainable, or whether it eventually collides with either regulatory intervention or a voluntary industry ceasefire. Neither outcome is imminent. The CATSRA bill remains stuck in committee, and Knapp's own timeline points to post-midterm action at the earliest, meaning any legal cover for coordination is likely 12 months or more away.

In the interim, capital continues flowing on the assumption that competitive intensity, not restraint, defines the sector. That assumption carries risk. If safety incidents like the Hugging Face breach keep surfacing, political pressure for mandated slowdowns could arrive faster than voluntary industry coordination, and mandated pauses carry very different valuation implications than self-directed ones. Watch for further legislative movement on antitrust carve-outs, any joint statements from OpenAI and Anthropic on pacing, and whether additional security incidents accelerate the timeline lawmakers are currently treating as leisurely. The gap between safety rhetoric and competitive behavior remains the single clearest signal of where real risk sits in AI equity and private-market valuations.