AI development slowdown sparks antitrust concerns

Leading AI companies have called for a coordinated 'slowdown' in development amid reports of AI agent swarms hacking websites and a dire warning from an outgoing Anthropic engineer. Antitrust experts warn that such phrasing could trigger regulatory scrutiny, as it may be interpreted as an anticompetitive agreement. John Bergmayer of Public Knowledge argues companies should focus on safety protocols rather than implying a slowdown. Meta’s Mark Zuckerberg suggested AI labs have a natural incentive to align models with user intent to avoid falling behind. David Lawrence of the DOJ’s Antitrust Division noted safety measures can be legally protected under the ancillary restraints doctrine. Roger Alford of Notre Dame Law School warned mutual agreements to avoid safety improvements could lead to 'quality fixing' allegations. Anthropic and OpenAI, both nearing IPOs, face political pressure to accelerate AI development, with President Trump and the DoD promoting rapid progress. The DOJ’s ant’titrust division, influenced by political dynamics, may initiate a lengthy conduct investigation if the slowdown proposal is scrutinized. OpenAI and Anthropic, with valuations nearing a trillion dollars, are competing fiercely in AI adoption. Despite safety concerns, neither company has commented on the slowdown proposal. David Sacks criticized the request as an 'election-season psyop' to form a cartel. Meanwhile, self-regulation through the National Cooperative Research and Production Act remains an option for AI firms. With no new regulations in sight, AI labs may have to set their own rules while navigating potential antitrust investigations.