Key points
- A Texas company says Anthropic's Claude models infringe five patents
- Anthropic recorded $450 million in authors' settlement liability at Sept. 30
- Anthropic won one Pentagon case and lost another
Anthropic (ANT) faces a new patent lawsuit from a Texas company that says its Claude AI models use inventions covered by five neural-network patents. The company filed the suit on Nov. 12, six days before Anthropic plans to price its IPO on Nov. 18.
The suit adds to the legal cases facing Anthropic as it prepares to go public. Its Nov. 10 IPO filing describes copyright suits in the US and Canada, a contract dispute with Reddit, and consumer class actions. It also describes an FTC information demand and Anthropic's own lawsuits against the Department of War.
This tracker uses the status Anthropic gave in the filing unless we cite a newer source. We'll add cases and rulings as they come.
A Texas company says Claude infringes five patents
Flinn Artificial Intelligence (AI) Technologies, based in Woodville, Texas, filed the suit in federal court in Delaware, according to the docket. Its patents name Steven D. Flinn and Naomi Felina Moneypenny as inventors. They filed applications in 2015 and 2016 describing neural networks that use multiple layers of attention, the complaint says.
Flinn asserts five patents against Claude models from Claude 1 through Claude 5, including all Opus, Sonnet, Haiku, Fable, and Mythos versions, as well as products such as Claude Code. It seeks at least a reasonable royalty in damages, an ongoing royalty after judgment, and up to three times the damages if the court finds the infringement was willful. The complaint doesn't ask the court to stop Anthropic from selling Claude.
A University of Tennessee foundation's patent case awaits a ruling
The University of Tennessee Research Foundation sued Anthropic in the same Delaware court on July 20, 2026, over two neural-network patents. Anthropic moved to dismiss the case on Oct. 13. Briefing finished on Oct. 28, and Anthropic asked for oral argument on Oct. 30, the docket shows. The docket showed no ruling from Judge Jennifer L. Hall as of Nov. 12.
Anthropic is still paying its $1.5 billion authors settlement
A group of authors sued Anthropic in August 2024 in federal court in Northern California, in a case called Bartz, alleging that it used copyrighted books to train its models. Anthropic agreed in September 2025 to settle for $1.5 billion, and the court approved the deal in July 2026, according to the filing. Anthropic paid $300 million in 2025 and still had $450 million recorded as a liability on Sept. 30, 2026.
Some authors and publishers opted out of the settlement and filed their own suits. Those cases are at various stages, the filing says.
Music publishers have filed a series of lyrics cases
Concord Music Group and other publishers sued Anthropic in October 2023, alleging that it copied and distributed copyrighted song lyrics. The case started in Tennessee and moved to Northern California. Both sides asked for summary judgment, and the court held a hearing on Oct. 21, 2026.
Between January and October 2026, publishers including Concord, Universal Music Publishing Group, Sony Music Publishing, and Warner Chappell filed more complaints in the same court. Some of them also name two Anthropic executives as defendants, the filing says. Those cases are at a preliminary stage.
Canadian authors and news publishers have sued in three provinces
Anthropic faces four copyright actions in Canada. Two proposed class actions on behalf of authors were filed in 2025, one in British Columbia and one in Quebec, and the Quebec case is on hold until the British Columbia case reaches a final judgment. Quebec news publisher La Presse sued in July 2026, and a group of Canadian news publishers sued in Ontario in October 2026.
Reddit's case is down to three claims
Reddit sued Anthropic in California state court in June 2025 over the use of Reddit data to build its models. In September 2026, the court narrowed the case to Reddit's claims for breach of contract, interference with contract, and unfair competition, according to the filing.
Subscribers have filed advertising and antitrust suits
Anthropic faces several proposed class actions alleging that it marketed Claude subscription plans with false or misleading advertising, the filing says. Those cases are generally at an early stage.
Four AI subscribers also sued Anthropic, OpenAI, SpaceXAI, and Google on Sept. 18 in federal court in Northern California. The proposed class action, Buist v. Anthropic, says the companies agreed to slow down AI development in violation of antitrust law.
Advocacy groups sued to stop new AI models
Four nonprofit groups sued Anthropic and four other AI developers, including Google, Meta, and OpenAI, on Nov. 12 in San Francisco Superior Court, The Associated Press reported. The plaintiffs are the Irreplaceable Education Fund, Public Citizen, Heat Initiative, and the Alliance for Secure AI. They ask a judge to stop the companies from developing or deploying new AI models, alleging violations of California's unfair competition law. The suit came after Anthropic's Nov. 10 filing, so the filing doesn't mention it.
Anthropic sued the Department of War and won in one court
Anthropic is the plaintiff in its fight with the Pentagon. It sued after the President directed federal agencies in February 2026 to stop using its technology and the Department of War designated it a supply-chain risk. A federal court in Northern California ruled largely in Anthropic's favor on Aug. 27, finding that the government's actions were retaliation for speech protected by the First Amendment. The government has appealed that ruling.
A separate case went the other way. On Sept. 25, a divided panel of the D.C. Circuit Court of Appeals upheld a second supply-chain risk designation, made under a different law, by a 2-1 vote. Anthropic has asked the full court to rehear the case, the filing says. We covered both rulings when Anthropic first filed for its IPO.
The FTC sent Anthropic a demand for information
The Federal Trade Commission issued a civil investigative demand to Anthropic in October 2026, seeking information about its consumer protection, data protection, and privacy practices and the safety and security of its products. The demand doesn't assert any claims, according to the filing. We reported on the FTC's demands to Anthropic and OpenAI in October.
Key dates
Coming up: Anthropic plans to price its IPO on Nov. 18. Rulings are pending on Anthropic's motion to dismiss the University of Tennessee case and on the summary judgment motions in the first Concord case. Anthropic hasn't yet responded to Flinn's complaint.
So far: Concord's first lyrics suit (October 2023); the Bartz authors suit (August 2024); Reddit's suit (June 2025); the Bartz settlement (September 2025); the federal ban and Pentagon designation (February and March 2026); the University of Tennessee patent suit (July 20, 2026); court approval of the Bartz settlement (July 2026); the Northern California ruling for Anthropic (Aug. 27, 2026); the Buist antitrust suit (Sept. 18, 2026); the D.C. Circuit ruling against Anthropic (Sept. 25, 2026); the FTC demand (October 2026); the Concord summary judgment hearing (Oct. 21, 2026); the advocacy groups' suit to stop new models (Nov. 12, 2026); and the Flinn patent suit (Nov. 12, 2026).
