Ex-DOJ Antitrust Chief Jonathan Kanter Pushes Back on AI Safety Exemption Talk

Jonathan Kanter, who led the Justice Department's antitrust division under Biden, says letting AI companies off the antitrust hook to chase safety goals would concentrate power rather than protect the public.
Jonathan Kanter, the Justice Department's antitrust chief for most of the Biden administration, used a Decoder interview published this week to reject the idea that artificial intelligence developers need relief from competition law in order to build safer systems. The argument, floated by some industry figures and investors, holds that splitting big AI labs or blocking their deals could slow work on existential risks. Kanter's answer, in short: competition and safety are not in tension, and treating them as rivals hands a handful of firms more leverage over the technology everyone else depends on.
Kanter ran the DOJ's Antitrust Division from November 2021 until he stepped down in December 2024, a stretch that included the government's monopolization suit against Google and its challenge to the JetBlue-Spirit merger. He now teaches law at Washington University in St. Louis and technology policy at Carnegie Mellon University, a pairing that puts him in front of both the legal and technical sides of the debate. His skepticism of the exemption idea lands as regulators in the US and Europe weigh how existing merger rules apply to model developers, cloud providers and chipmakers.
The exemption argument has circulated for at least two years, often attached to warnings that frontier models could escape human control. Its critics point out that the companies making the loudest safety claims are also the ones best positioned to benefit from a lighter regulatory touch. Kanter's framing puts the burden on that logic: if a capability is genuinely dangerous, the fix is oversight of how it is built and deployed, not permission for a few firms to grow unchecked.
Antitrust enforcers have already signaled they will look closely at AI partnerships rather than only outright acquisitions. The Federal Trade Commission opened a study of investments by Alphabet, Amazon and Microsoft into AI startups, and the DOJ has examined arrangements between model developers and chip suppliers. Those inquiries treat minority stakes and compute deals as potential ways to extend control without a formal merger, a theory that an exemption would directly undercut.
Kanter's comments arrive as the Trump administration's antitrust posture remains unsettled, with leadership changes at both agencies and several pending cases against large technology companies. What happens to those suits will shape whether AI firms face real limits on acquisitions or get the looser treatment some executives have sought. The open question is whether lawmakers take up AI competition as standalone legislation or leave it to courts and agencies to define case by case.