AI Firms' 'Slowdown' Wording Could Invite Regulatory Scrutiny
Leading AI companies have called for a coordinated development slowdown after reports of rogue agent swarms and a warning from an outgoing engineer, but their phrasing may trigger antitrust concerns. Legal experts note that using terms like 'slowdown' can imply an anticompetitive agreement to reduce output, even if the intent is safety-focused. The companies could have avoided this by framing their efforts as security improvements rather than a collective pause.
The article highlights how antitrust scrutiny often focuses on the language companies use as much as their actual conduct. Google reportedly trained employees to avoid phrasing that could imply anticompetitive behavior, instead emphasizing consumer benefits. Legal experts note that a collective "slowdown" without clear purpose could be read as an agreement to reduce output, whereas framing efforts as joint safety protocol work would likely avoid that interpretation.
The piece also notes that some legal voices see safety coordination as compatible with competition law. David Lawrence, formerly of the DOJ's Antitrust Division, argued that preventing catastrophic risks actually promotes competition and is protected under the ancillary restraints doctrine. One FTC attorney added that "no humanity would result in no competition," underscoring the tension between safety coordination and antitrust enforcement.
The way AI companies frame their safety coordination could shape regulatory outcomes for years. If regulators interpret "slowdown" as collusion, firms may face costly investigations that distract from actual safety work. Conversely, if the framing is accepted as legitimate safety coordination, it could set a precedent for how the industry handles future risks. The public may see either outcome as affecting how quickly AI products reach them and how much oversight the industry receives.