The Trump administration on October 7 publicly branded frontier AI labs’ bid for a federal antitrust exemption as “moat-digging,” according to Law360, the sharpest White House rebuke yet of a proposal that would let the largest model-makers legally coordinate a slowdown of their own industry.

The vehicle is the Schiff-Banks-Whitesides-Latta bill, which per CNBC would cover coordinated “delaying or limiting the deployment of high-risk models.” The intellectual scaffolding is Anthropic CEO Dario Amodei’s “We Must Pace the Frontier” essay, which asked Washington to “issue a narrow waiver for certain kinds of safety conversations.” OpenAI’s Sam Altman and Google DeepMind’s Demis Hassabis endorsed it.

The pushback is bipartisan and unusually blunt. Jonathan Kanter, the former DOJ antitrust chief, told CNBC that nothing in current law blocks labs from sharing safety information; a waiver, he argued, is only required to agree to “stop innovating and stop competing.” Sen. Elizabeth Warren was shorter: “No… big companies love antitrust waivers, but no.”

The posture is reinforced by live enforcement. The FTC’s industry-wide probe into Anthropic, OpenAI, and other labs is ongoing, and four paying AI subscribers have filed a proposed class action alleging Anthropic, OpenAI, Google, and xAI already coordinated a slowdown illegally.

For small businesses, the stakes are concrete. Open-source models are, per Washington Monthly, “often far cheaper to use than frontier offerings,” and that pricing pressure is what keeps the frontier honest. A government-blessed pact to pace deployment would blunt it. See also our coverage of Congress leaving town with no federal AI law until 2027 and Trump’s new AI Force and incoming AI czar.

Congress is out until after the midterms. The waiver fight resurfaces the moment it returns, into an administration that has already named the bid for what it’s.

Sources

Sources