David Sacks tells Anthropic and OpenAI: pace the frontier without permission

Former White House AI Czar David Sacks posted that if Anthropic’s Dario Amodei and OpenAI’s Sam Altman want to pace the frontier, they should do it themselves and stop pretending they need outsiders’ permission, antitrust carve-outs, or a preferred regulatory framework.

The former White House AI lead put a money-and-power frame on the slowdown debate. His argument is the news. A lab slowdown or a waived antitrust rule is not.

Former White House AI Czar David Sacks posted a long note on X answering Anthropic CEO Dario Amodei’s “pace the frontier” line and OpenAI CEO Sam Altman’s agreement. The note is timestamped 03:14 UTC. Sacks’s response, in his words: “go ahead.” That note is the filing event. This is Sacks’s public argument. It is not a White House order, not a statute, and not a lab product change. This filing is Sacks’s reply only.

Sacks writes that Amodei and Altman “are the frontier.” By “any reasonable metric — market share, revenue growth, model capability,” he says the two have “a duopoly on frontier intelligence,” and that they have also “claimed the lead is widening because of recursive self-improvement.” Those are his characterizations — not a desk market study.

He says he does not see what they see in the lab. If the unreleased models are “scary enough” that they think they should slow down, he supports their “decision to be responsible.” The ask is that they slow themselves.

Quoted line: “But stop pretending you need anyone else’s permission.” The same paragraph tells them to stop pretending antitrust law has to be suspended so they can form a cartel; stop pretending they need a regulatory approval process that supersedes product liability; stop pretending METR is independent when it is “intertwined with Anthropic’s investors and staff”; and stop pretending those same evaluators need to police competitors “who aren’t even at the frontier.” Those charges are Sacks’s.

He writes that the motivation to slow down is not “purely altruistic.” In his account they face “massive product-liability exposure” if their products enable a truly damaging cyberattack; the market already punishes unpredictable or unauthorized model behavior; and after the Hugging Face episode it is “simply good business” for OpenAI and Anthropic to trade some raw power for reliability and predictability. “Call it alignment if you want. It is also just giving customers what they want.” Treat that as his business reading of the OAI-HF episode — not a new incident report.

On geopolitics and rules: pacing, he writes, would create breathing room for a more intelligent conversation about regulation than Bernie Sanders’ “shut it all down.” China, he adds, is “very unlikely to join a global agreement,” and that has to be taken into account. He says they are the ones setting the frontier, and “the easiest way not to build superintelligence is for you to agree not to build it.” Demanding their preferred regulatory framework as the price of that, he writes, “will look like blackmail of the public and the political system.”

Close of the note: “So just do it. If you do, you’ll buy goodwill for the next conversation. If you don’t, we’ll know this was just another bid for regulatory capture — or an election-season psyop.” CONFIRMED here: the posted text.