Does AI need an antitrust exemption so it doesn’t kill everyone????

In this Decoder interview, former DOJ antitrust chief Jonathan Kanter discusses the debate over AI safety and regulation, focusing on whether frontier AI companies should receive an antitrust exemption to coordinate on safety. He offers both a generous interpretation—that companies genuinely fear the pace of innovation and lack of regulation—and a cynical one—that they seek to slow competition to manage spending and IPO valuations. Kanter argues that existing antitrust laws already permit legitimate safety collaboration, such as threat-sharing clearinghouses, without exemption. He compares AI to products liability law, citing the Meta case over child safety, and says companies should be liable when their AI agents cause harm. He criticizes Congress for inaction and the Trump administration's hands-off approach, while noting states have done better. On China, he calls the threat largely a "convenient boogeyman," arguing the U.S. wins by promoting domestic competition and safe products, not by tolerating monopolies. He warns against a "national champions" approach and defends his antitrust legacy, noting wins against Google and Ticketmaster. He concludes with prescriptions: clear liability for harmful products and congressional enumeration of societal values like mental health and competition.

(Source：The Verge)