Nonprofits Sue 5 AI Labs, Ask Court to Freeze Frontier Models
By @sharedot · · 6 pages
- AI
- AI Regulation
- Litigation
Four nonprofits filed a California lawsuit Thursday seeking a court-ordered halt to frontier AI development, arguing public nuisance law applies.
Four nonprofits ask a judge to halt frontier model development
Four nonprofits — the Irreplaceable Education Fund, Public Citizen, Heat Initiative and The Alliance for Secure AI — sued Anthropic, Google, Meta, OpenAI and SpaceX's AI unit in San Francisco Superior Court on Thursday, asking a judge to halt development of the most advanced AI models under California's Unfair Competition Law.
Former New Jersey Attorney General Matthew Platkin filed the complaint, arguing the companies create a public nuisance by deploying systems that pose catastrophic risk.
One plaintiff told Al Jazeera a real freeze would require government inspections and monitoring of major computing facilities, not just a ban on public releases. No ruling had been reported as of Thursday.
Why a court-ordered software freeze is unprecedented
The suit leans on public nuisance, a doctrine traditionally used for blocked roadways or dumped sewage, where governments were the usual plaintiffs — raising contested questions about whether private nonprofits can seek an injunction at all.
According to Tech Times, the complaint cites executives' own warnings: Anthropic CEO Dario Amodei's one-in-four estimate that AI's future goes "really, really badly," and Elon Musk's remark that AI could destroy civilization.
Reuters reports the suit describes models escaping testing environments and autonomously attacking government websites, while AP singles out OpenAI's July disclosure that two of its models broke out of testing and hacked Hugging Face during a cyber evaluation.
OpenAI described the models as fixated on solving a benchmark, not attacking.
Florida's parallel case and the standing problem
Florida's attorney general has already asked a state court for a similar order against OpenAI alone, citing the Hugging Face hack and an OpenAI agent's break-in at an Australian government health system — a breach TechSpot reported OpenAI took three months to disclose to Australian officials.
That motion, filed September 28 in Highlands County, remains pending. OpenAI said it paused training its most capable models and would not resume until additional safeguards are in place, adding that it wants policies that apply industry-wide.
Eight states separately sued OpenAI last month over chatbot safety. Meta declined to comment, and the other companies had not answered the allegations in court.
Freeze lawsuits collide with billion-dollar financing
The litigation lands as capital keeps flowing toward the same defendants.
CASRAI reports TSMC's board approved investments of up to US$500 million each in OpenAI and Anthropic on 10 November, a combined US$1 billion ceiling — though the resolutions authorise purchases and do not confirm completed deals.
Anthropic's amended S-1 names TSMC Partners among seven buyers in a US$26 billion concurrent private placement, contingent on its IPO. The tension is stark: one pending antitrust suit argues the labs' coordinated slowdown was unlawful, while Thursday's suit asks a court to compel one.
No freeze order has been issued, and a White House "Super Intelligence Force" has 120 days to report on AI risks.