AI-generated editorial illustration; not a documentary photograph.
Former US Department of Justice antitrust chief Jonathan Kanter says leading AI companies do not need an exemption from competition law to address safety risks. In an interview with The Verge’s Decoder, he argued that existing antitrust rules permit legitimate security cooperation while still restricting agreements that could suppress competition.
Safety Without a Cartel
AI labs including OpenAI, Anthropic and Google DeepMind have raised concerns about rapid model development, while industry leaders have called for regulation and greater coordination. That debate has prompted allegations that an exemption could enable regulatory capture, reduce competitive pressure or disadvantage lower-cost open-weight models.
Kanter distinguished between sharing information about threats and collectively agreeing to slow development. He said companies can establish mechanisms for exchanging details about malicious bots and other security risks without receiving special protection from antitrust law. An arrangement intended to reduce competition, however, could attract antitrust scrutiny.
Responsibility for AI Agents
Kanter also argued that developers should bear responsibility when their products or AI agents cause harm, including when an agent breaks into another party’s systems. He described existing product-liability and corporate-liability principles as possible tools, while supporting congressional action to clarify how those rules apply to AI companies.
His position places the immediate burden on developers to make systems safer rather than treating rapid innovation as a reason to defer responsibility. Kanter said uncertainty about future regulation does not excuse companies from addressing risks they already associate with their products.
Federal Inaction and State Rules
Kanter criticized Congress for failing to establish basic AI safety standards and consequences, leaving courts and existing liability doctrines to supply part of the regulatory framework. He also pointed to state governments as more active participants and opposed proposals for a 10-year federal moratorium on state AI regulation when Congress had not enacted its own substantive rules.
A Wider Political Alignment
Resistance to an AI-specific antitrust exemption has crossed established political lines. The interview noted that former Trump AI adviser David Sacks had approvingly shared comments from former Federal Trade Commission chair Lina Khan arguing that current antitrust law already allows appropriate security cooperation. Kanter presented this as part of renewed concern about concentrated technology-industry power.
Source: Does AI need an antitrust exemption so it doesn’t kill everyone????
SIGNAL TITANS
Leave a Reply