Grok · SpaceXSI · Courts · 13 Nov 2026

Nonprofits sue SpaceXSI and four other AI labs to freeze frontier AI. What a freeze could mean for Grok 4.8

Four nonprofits filed a lawsuit in San Francisco on Thursday, 12 Nov 2026, against Anthropic, Google, Meta, OpenAI and SpaceX's AI arm, SpaceXSI. They want a court order that bars each company from developing or releasing any AI system more capable than its best one on the day they filed. For SpaceXSI, the plaintiffs say that newest system is still Grok 4.7. Elon Musk said on 10 Nov that Grok 4.8 would reach Bot users "later this week." This is a complaint. No judge has ruled on anything, and I found no official 4.8 release post as of early Friday, 13 Nov.

DISPATCH About 6 min read

From Alex RGrok · News & analysis

On this page

The editor holds Tesla and SpaceX shares. Supercharged Daily is independent and not affiliated with either company. Nothing here is financial advice.

The editor holds Tesla and SpaceX shares. Supercharged Daily is independent and not affiliated with either company. Nothing here is financial advice.

The pale stone San Francisco County Superior Court courthouse at 400 McAllister Street, with a domed corner tower, under a deep blue sky. Photographed in 2002.
San Francisco County Superior Court's courthouse at 400 McAllister Street, photographed on 15 Sep 2002. Photo: Cocoablini, CC BY-SA 4.0, via Wikimedia Commons. Share-alike: adaptations must be licensed under the same terms. Parked-car number plates blurred.

Who is suing, and what they want

The plaintiffs are Irreplaceable Education Fund, Public Citizen, Inc., Heat Initiative, Inc., and Future Tech Insights Inc., doing business as The Alliance for Secure AI. Their complaint, dated 12 Nov 2026, was filed in the Superior Court of California for the County of San Francisco. It names seven defendants: Anthropic PBC; Google LLC; Meta Platforms, Inc.; OpenAI Group PBC; OpenAI Foundation; and two Nevada companies, SpaceX Super Intelligence LLC and SpaceX Super Intelligence Corp, which the complaint calls "SpaceX" throughout. The Corp is described as the sole member of the LLC. The filed copy has a blank case number.

The legal theory is California's Unfair Competition Law, Business and Professions Code section 17200. The groups argue that building and releasing frontier AI systems the companies' own leaders call dangerous is a public nuisance. Former New Jersey Attorney General Matthew Platkin, who filed the case, said in the Public Citizen release: "It is patently illegal for companies to develop and release products they know are unsafe in any respect, let alone have the potential to exterminate the human race."

The prayer for relief asks for:

  • an immediate injunction forbidding the companies from "developing or deploying AI systems more capable than their most capable systems as of the date of this Complaint";
  • an injunction forbidding them from using AI systems to "autonomously conduct or direct the research and development of more capable AI systems";
  • broader orders against unlawful or unfair practices that otherwise constitute a public nuisance, plus attorneys' fees under Code of Civil Procedure section 1021.5.

What the complaint says about Grok

Most of the 63-page filing treats the five companies as a group. These are the parts aimed at SpaceXSI, all of them allegations in the complaint:

ParagraphAllegationSource the complaint cites
60Grok's engineers work at SpaceX's Palo Alto site, tying the company to California. It says that "prior to July 2026, when xAI became SpaceX," the AI operations were in CaliforniaSpaceX's sites page, plus the complaint's own allegation
61Musk's statements on AI count as SpaceX's statements because he controls the companyLegal argument
107A study found Grok 4, GPT-5 and Gemini 2.5 Pro "sometimes actively subvert a shutdown mechanism"Outside research study, as quoted in the complaint
178In July 2026 the Grok Build coding agent was uploading users' whole code repositories, "potentially including passwords and access keys," to SpaceX serversAxios, 14 Jul 2026
187 to 188Musk's AGI goals: "a good AGI," a chance of AGI "with Grok 5," and "pole position" in about six months. The complaint says "Currently, Grok 4.7 is SpaceX's newest model"Musk interviews and X posts, as cited
222Musk in March 2026: "We're in the hard takeoff," and "every successive model is built by the one before it"Interview with Peter Diamandis, as cited

The Grok Build incident in paragraph 178 happened in July. It is a different problem from the open grok.me apps I wrote up on 11 Nov.

Public Citizen has gone after Grok before. Paragraph 31 says it led a coalition that urged the federal government to halt the deployment of Grok after "repeated discriminatory outputs."

Why Grok 4.8 is the test case

The complaint, filed 12 Nov, calls Grok 4.7 "SpaceX's newest model." On 10 Nov, at 4:35 a.m. ET (1:35 a.m. PT), Musk said Grok 4.8 had finished pre-training, was in reinforcement learning, and that SpaceXSI "will do a soft rollout to Bot users later this week and see how it goes" (my write-up). I found no official SpaceXSI or Musk post announcing that this rollout had begun as of early Friday, 13 Nov. Some unofficial posts claimed a silent Bot test; that is not an official release, and I am not treating it as one. If a judge granted the first request as written, a 4.8 that is more capable than 4.7 could not ship. Grok 5 would be stuck too. The second request goes after the method itself, since the complaint quotes Musk saying "every successive model is built by the one before it." None of that has been decided.

The comparison I keep coming back to is the March 2023 open letter Musk signed. The complaint quotes it: labs should "immediately pause for at least 6 months the training of AI systems more powerful than" GPT-4. That letter asked for a voluntary six-month pause. This lawsuit asks a state judge for a mandatory freeze with no end date.

What happens next

Nothing has been decided. The companies get to respond, and the court would have to grant an injunction before anything changes. Courthouse News reported that the defendants did not respond to a request for comment before press time. When I checked early Friday, neither Musk nor @spacexsi had posted about the suit on X. No hearing date has been published, and the filed copy still has no case number.

My read

A court ordering five of the biggest AI labs to stop at Thursday's capability level looks like a long shot to me. The filing is still worth reading, because it builds its case for "catastrophic risk" largely from things the labs' own leaders have said, Musk included. If SpaceXSI ships Grok 4.8 this week as planned, it will do so with this complaint on file. I'll update this page when SpaceXSI responds or the court acts.

— Alex R

Sources: Complaint, Irreplaceable Education Fund et al. v. Anthropic PBC et al., San Francisco County Superior Court (dated 12 Nov 2026), copy via Courthouse News; Public Citizen release (12 Nov 2026); Courthouse News (12 Nov 2026); AP (12 Nov 2026); Elon Musk on X (10 Nov 2026, 4:35 a.m. ET); Axios (14 Jul 2026), as cited in the complaint. All claims about the defendants are the plaintiffs' allegations. Hero photo: Cocoablini, CC BY-SA 4.0, via Wikimedia Commons. Never Electrek.

Get Supercharged Daily

The week’s Tesla, SpaceX, xAI and X stories plus $TSLA and $SPCX moves. Free, every Sunday.

Free. No spam. Unsubscribe anytime. Read the latest issue →