The race to build frontier artificial intelligence models has officially collided with antitrust reality. OpenAI has approached members of the U.S. Congress to determine whether an industry-wide agreement to coordinate and decelerate the deployment of advanced AI systems would breach federal competition statutes. This legislative outreach follows a public argument from OpenAI’s research leadership advocating for collective industry restraint.
Jakub Pachocki, chief scientist at OpenAI, argued in a public post that managing self-improving AI safely demands collaborative pacing, suggesting that voluntary multi-lab slowdowns must become standard practice until shared safety baselines are codified across the sector.
"coordinating to slow down future development"
In Pachocki’s framework, coordinating development schedules serves as a necessary risk control for autonomous and self-improving systems.
Antitrust Barriers and Legislative Stalls
Under U.S. competition law, however, noble intentions do not sanitize market collusion. Orchestrated agreements among dominant commercial competitors to delay product releases or throttle capability progress look uncomfortably like output restriction under Section 1 of the Sherman Antitrust Act. Nicholas Felstead, assistant director at the Australian Competition and Consumer Commission and former AI policy fellow at the Center for Law & AI Risk, noted in a March paper that an explicit consensus to pause or pace model development creates immediate antitrust exposure, warning that unresolved legal uncertainty functions as an aggressive deterrent.
Efforts to carve out a statutory safe harbor have already stalled on Capitol Hill. In July, a bipartisan group introduced the Collaboration on Adversarial Threats and Security Risks Act to shield AI labs from antitrust penalties when collaborating on security and systemic risks. The bill remains buried in the House Judiciary Committee without a markup date.
Caleb Knapp, director of government affairs at the AI Policy Network, indicated that while congressional interest in AI risk governance is growing, meaningful legislative intervention will likely remain on hold until after upcoming election cycles.
For enterprise buyers and executive decision-makers, this regulatory impasse has concrete commercial consequences. Institutionalizing collaborative pause mechanisms effectively cements incumbent dominance by freezing market positions under the banner of risk management. Until Congress clarifies the boundary between cartel behavior and responsible safety coordination, enterprise leaders face prolonged uncertainty that complicates multi-year AI capital allocation and infrastructure roadmaps.