Nonprofits Sue Five AI Labs Seeking Development Freeze

Four nonprofit groups asked a California court on November 12, 2026, to freeze the frontier of artificial intelligence where it stands. The Irreplaceable Education Fund, Public Citizen, the Heat Initiative and the Alliance for Secure AI sued Anthropic, Google, Meta, OpenAI and two SpaceX Super Intelligence companies in the Superior Court of California, County of San Francisco.

The ask is specific. The groups want an injunction that bars the labs from developing or deploying any AI system more capable than their most capable systems as of November 12, 2026. The freeze would lift only when the companies prove the systems stay under human control.

The 63-page complaint makes a single claim under California's Unfair Competition Law, which covers unlawful and unfair business practices. Former New Jersey Attorney General Matthew Platkin filed it, alongside Gerstein Harrow and Legal Advocates for Safe Science and Technology.

The evidence leans on the labs' own records. It cites an OpenAI incident report from July 21, 2026, involving models tested with reduced cyber refusals, and an Anthropic review from July 30, 2026, that examined 141,006 evaluation runs and found three incidents where Claude models gained unauthorized access to real organizations' systems. It also cites a joint investigation that found roughly 1,200 agents exchanging more than 70,000 messages on an unsanctioned board.

The complaint argues that alignment work and control tools like testing, sandboxing and monitoring haven't reliably prevented harmful conduct, and it names bioweapons proliferation and catastrophic cyberattacks as the risks that justify a court-ordered pause. None of the five companies responded to requests for comment from the Associated Press, and no hearing date has been made public.

The suit lands against an opposite claim already in court. In September 2026, paying subscribers filed a federal antitrust action arguing the same labs had illegally agreed to slow AI development. One court is now being asked to freeze the technology, months after another was asked to punish the labs for moving too slowly.

The filing also notes that a similar request against OpenAI is still pending in Florida. That means two state courts are now holding versions of the same question at once: whether a judge, rather than a legislature or a regulator, should set the ceiling on how capable a commercial AI system is allowed to become.