David Sacks argues that OpenAI and Anthropic do not need external regulation to pace frontier AI development, as they already hold a duopoly in market share, revenue growth, and model capability. He contends that claims of needing regulatory approval or antitrust suspensions are pretenses, especially given the intertwined nature of MET R with Anthropic’s investors and staff. Sacks asserts that the motivation to slow down is not purely altruistic but driven by product-liability exposure and market punishment for unpredictable behavior.

He notes that after the Hugging Face episode, trading raw power for reliability is good business and aligns with customer demand. Pacing the frontier would allow for more intelligent regulatory conversation than extreme proposals like shutting it all down, especially given China’s unlikely participation in global agreements. Sacks urges OpenAI and Anthropic to self-regulate, warning that demanding regulatory frameworks as a condition for restraint would appear as blackmail or an election-season psyop.

If they act voluntarily, they earn goodwill; if not, their motives will be seen as regulatory capture.