Trump’s undefined ‘AI Force’ meets an antitrust lawsuit over slowing the labs

President Trump announced an “AI Force” and said he would appoint a new AI czar, but supplied no command structure, budget, legal authority or relationship to existing regulators. He framed the project as an instrument of acceleration: the US must beat China, AI could eventually contribute as much as 25% of GDP, and fears about uncontrolled systems or data centres are, in his telling, politically manufactured obstacles. The czar position has been vacant since March.

The institutional problem arrived immediately. A proposed class action filed in northern California accuses Anthropic, OpenAI, SpaceXAI and Google of illegally agreeing to slow AI development after their leaders publicly endorsed parts of Dario Amodei’s September 12 call for coordinated restraint. The named plaintiffs pay for ChatGPT, Claude, Grok or Gemini and say less competition would reduce the value of their subscriptions. The companies had not responded when the Associated Press reported the suit.

An endorsement of safer development is not itself proof of a binding cartel, and the complaint must still establish an actual agreement and consumer harm. But the case exposes the missing governance mechanism. Labs can each slow unilaterally; collective pacing can be more effective, yet also resembles coordination among dominant competitors. Amodei anticipated that tension and suggested a narrow government waiver for safety talks. Trump’s announcement does not say whether the AI Force would mediate such talks, set minimum standards, investigate incidents or merely promote deployment. Until those powers are specified, “AI Force” is a slogan where an institution is needed.