AI Slowdown Proposal Triggers Antitrust Lawsuit Against Four AI Companies

Summary

A proposed class action in the U.S. District Court for the Northern District of California accuses Anthropic, OpenAI, Google, and SpaceXAI of illegally coordinating to restrain AI development. The suit was filed after Anthropic CEO Dario Amodei called for an industry-wide pause of roughly one to two years of frontier capability gains so safety work could catch up with risks such as recursive self-improvement. Sam Altman, Elon Musk, and Demis Hassabis expressed support in principle, while the Trump administration, Meta CEO Mark Zuckerberg, NVIDIA CEO Jensen Huang, European competitors, and Senator Josh Hawley opposed the proposal or warned that it could entrench incumbent firms. The article links the dispute to earlier calls for shared safety protocols, including a July statement signed by more than 1,100 employees from OpenAI, Anthropic, Google DeepMind, and Meta; signatures later exceeded 1,300. It also notes that Anthropic was reportedly considering a faster model release after OpenAI released GPT-6 Astra, a move that would conflict with the public case for restraint. The proposed slowdown raises particular concerns in Europe, where firms such as Mistral fear coordinated U.S. release strategies and regulation of open-source AI could limit competitors that have fewer resources. The article further examines the financial incentives around the debate: Anthropic is reportedly pursuing a possible IPO valued near $2 trillion, while NVIDIA was said to be considering up to $10 billion as an anchor investment. It describes Jaan Tallinn’s simultaneous support for AI-safety organizations and role as an Anthropic Series A leader and board observer. The White House has rejected a coordinated slowdown, and Hawley rejected an antitrust waiver for safety cooperation. The lawsuit could therefore use discovery to establish what companies agreed to, when they agreed, and whether the restraint was substantive or merely performative.