AI companies increasingly agree that frontier development may need to slow down. The harder question is whether they are legally allowed to coordinate that slowdown.
OpenAI has approached members of Congress to determine whether coordinating an industry-wide slowdown on frontier artificial intelligence models would trigger antitrust violations, according to reports first published by WIRED.
The issue creates an unusual policy conflict: individual companies can slow development on their own, but competitors coordinating limits on output or product releases could invite federal scrutiny even when the stated goal is AI safety.
Under the Sherman Antitrust Act, formal agreements between competing corporations to throttle product rollouts can be interpreted as illegal output restrictions. While a single lab is free to unilaterally halt operations, mutual pledges between fierce rivals invite federal scrutiny.
That uncertainty may itself discourage companies from pursuing collective safety measures. “Legal uncertainty can act as a powerful deterrent,” Nicholas Felstead, an assistant director at the Australian Competition and Consumer Commission, wrote in a March paper examining coordinated AI s, according to WIRED.
Lawmakers have introduced a potential shield: the bipartisan Collaboration on Adversarial Threats and Security Risks Act, sponsored by Sens. Adam Schiff and Jim Banks. The proposal would permit safety collaborations subject to advance notice to the Department of Justice.
However, the measure remains in the House Judiciary Committee, leaving executives without a legal safe harbor ahead of the upcoming midterm elections.
Industry leaders agree on the risk, not the solution #
The outreach comes amid rare public alignment among major frontier AI leaders. Anthropic CEO Dario Amodei publicly warned on Saturday that companies must “slow the pace at which we improve the capabilities of AI models.” OpenAI CEO Sam Altman quickly endorsed the sentiment, posting on X: “I agree with Dario that we need to pace the frontier.” xAI chief Elon Musk echoed the stance, writing, “Dario is right.”
Yet skepticism remains widespread across both industry and government. Rep. Josh Gottheimer, D-N.J., dismissed the self-policing calls as hollow. “If Dario and Elon are truly worried, they can pump the brakes at their own labs, today,” Gottheimer told POLITICO.
John Schulman, an OpenAI co-founder now serving as chief scientist at Thinking Machines, similarly argued that antitrust worries are being weaponized as an excuse for corporate hesitation. “They’ll cite antitrust, but that’s fake—antitrust prohibits certain agreements, but not from jointly developing a proposal,” Schulman wrote on X.
More must-read AI coverage
-
[SS&C Intralinks DealCentre AI vs. Datasite: Which platform is built for the future of dealmaking?](https://www.techrepublic.com/article/dealcentre-ai-vs-datasite/) -
[SS&C Intralinks FundCentre AI vs. Juniper Square: Which platform better supports modern private markets fund managers?](https://www.techrepublic.com/article/fundcentre-ai-vs-juniper-square/) -
[Why Data, Not Models, Determines AI Success](https://www.techrepublic.com/sponsored/why-data-not-models-determines-ai-success/)
[The Rise of the AI-Native Factory: How Physical AI Is Transforming Manufacturing](https://www.eweek.com/a/artificial-intelligence/the-rise-of-the-ai-native-factory-how-physical-ai-is-transforming-manufacturing/)
The entrenchment trap #
By lobbying Washington to define the legal boundaries of a market-wide , frontier developers are shifting attention away from unilateral accountability.
If reigning pioneers construct a government-approved coordination framework, they risk locking in a dominant cartel: existing incumbents preserve market share at current capability thresholds, while prospective competitors face bureaucratic friction to advance further. Giving major developers greater freedom to jointly slow certain forms of AI development could reduce the antitrust risk surrounding collective safety measures. But policymakers would also have to prevent those arrangements from becoming barriers that protect established companies from newer competitors.
That is the difficult balance behind OpenAI’s request. Too little coordination could leave companies reluctant to cooperate on shared risks. Too much protection could give the largest labs greater influence over how quickly the market is allowed to advance.
For enterprises adopting advanced AI systems, however, any industry-wide restraint remains hypothetical. Companies still need to assume that frontier models will continue improving rapidly and build their own controls around how those systems are deployed. In other AI news: President Donald Trump dismissed warnings that AI could cause human extinction**, saying his bigger concern is ensuring the U.S. stays ahead of China as the race to develop increasingly powerful systems accelerates.**