AI 'Slowdown' Rhetoric Creates Antitrust Vulnerability

Leading AI companies have called for a coordinated development "slowdown" following reports of AI agent swarms hacking websites and warnings from a former Anthropic engineer. However, as Wired first reported, the specific language used to frame this effort may invite regulatory scrutiny under the Sherman Act.

John Bergmayer, legal counsel for Public Knowledge, notes that antitrust economists typically examine whether companies are reducing output. Bergmayer argues that by phrasing their efforts as a "slowdown" or "pause" rather than focusing on the development of safety protocols, AI labs may have "boxed themselves into a corner," as such terms can be interpreted as anticompetitive agreements to reduce trade.

Alternative framing could have mitigated these risks. Meta CEO Mark Zuckerberg declined to endorse an explicit slowdown, arguing instead that companies have a "strong natural incentive" to solve "misalignment" because those who fail to make agents behave better will fall behind competitively. Similarly, David Lawrence, a former policy director for the Department of Justice’s Antitrust Division, stated on LinkedIn that agreements to prevent catastrophic risks are protected under the "ancillary restraints doctrine" because they promote competition and increase output.

Despite these legal protections, some critics view the push for exemptions as a tactical move. David Sacks, cochair of the President’s Council of Advisors on Science & Technology, accused OpenAI and Anthropic of acting as a duopoly and described the request for an antitrust exemption as an "election-season psyop" designed to form a cartel. Conversely, Roger Alford, a Notre Dame Law School professor and former DOJ Antitrust Division official, suggests that failing to implement safety measures could lead to "quality fixing" allegations, citing a European case where automakers were fined roughly one billion dollars for agreeing not to compete on emissions-reducing technology.