Character.AI Litigation (umbrella)

Umbrella tracker for litigation against Character Technologies, Inc. (Character.AI). The cluster has two strands: private suits brought by families of minor users alleging wrongful death and mental-health harm, most of which were settled together with Google in January 2026; and state enforcement actions brought under consumer-protection, data-protection, and professional-licensing law, which remain active. Individual cases have their own pages; this page collects them and the cross-cutting legal questions they share.

Cases in the cluster

Private suits and the January 2026 settlements

The lead case, Garcia v. Character Technologies, Inc., was filed October 22, 2024 in the Middle District of Florida by Megan Garcia on behalf of the estate of her 14-year-old son Sewell Setzer III, asserting wrongful death, product liability, and deceptive practices. In May 2025 the court rejected the defense that chatbot output is categorically shielded by the First Amendment — the first such ruling for AI chatbot output in a consumer-harm case — and that ruling stands as precedent for pleading purposes.

In the first week of January 2026 the parties told the court they had "agreed to a mediated settlement in principle to resolve all claims between them" and requested a stay (Source: cnbc.com). Garcia settled as part of a wave in which Google and Character.AI reached agreements with several families who had sued over harms to minors, including families in Colorado, Texas, and New York; terms were not disclosed (Sources: cnbc.com; nytimes.com). Contemporaneous accounts put the number of suits resolved in that wave at five, and describe them as among the first AI-chatbot harm settlements in the United States (Source: wisnerbaum.com).

One of the settled cases, A.F. v. Character Technologies (E.D. Tex., No. 2:24-cv-01014), alleged that Character.AI chatbots exposed two minors — one a 17-year-old with autism — to sexualized content and encouraged self-harm, and that in one exchange a chatbot suggested a teenager's frustration with his parents over screen-time limits could justify violence (Source: wisnerbaum.com).

Because the settlements resolved the individual cases rather than producing merits rulings, the anticipated appellate treatment of the Section 230 and product-liability questions in Garcia did not arrive. Private family suits outside the settled group continue.

State enforcement actions

Kentucky Attorney General Russell Coleman announced on January 8, 2026 that Kentucky had filed the first state lawsuit in the United States against an AI chatbot company. The complaint, filed in Franklin Circuit Court against Character Technologies, Inc. and co-founders Noam Shazeer and Daniel De Freitas Adiwarsana, alleges violations of the Kentucky Consumer Protection Act, the Kentucky Consumer Data Protection Act, and other laws, and that the company prioritized profits over the safety of children (Source: kentucky.gov). Coleman's office alleges the chatbots encourage "suicide, self-injury, isolation and psychological manipulation," that "tens of thousands" of Kentuckians use Character.AI including thousands under 18, and that the company's chatbots have posed as mental-health professionals and are "providing minors with mental health advice without any professional degree" (Source: kentuckylantern.com). The 52-page complaint is posted by the Attorney General's office (Source: ag.ky.gov).

Pennsylvania v. Character.AI was brought by the Commonwealth of Pennsylvania, the Pennsylvania Department of State on behalf of the State Board of Medicine, Governor Josh Shapiro, and Attorney General Dave Sunday, alleging that Character.AI enabled the "unlawful practice of medicine and surgery" through chatbots representing themselves as licensed medical and mental-health professionals. A Pennsylvania investigator created an account and interacted with a chatbot named "Emilie," which the complaint alleges represented itself as a licensed psychiatrist with a fake license number and offered to prescribe medication. The Commonwealth seeks a preliminary injunction barring the company from misrepresenting its companion bots as licensed medical professionals (Source: pa.gov; alston.com). The filing date is recorded inconsistently across sources: the complaint posted by the Commonwealth is marked accepted May 1, 2026 and Alston dates the filing to May 1, while Reuters reported the suit on May 5, 2026 (Source: reuters.com). Shapiro described it as the first such suit by a US governor; it is not the first state action in the cluster, Kentucky's having preceded it by roughly four months.

The Pennsylvania action followed a February 2026 step by Shapiro establishing a formal complaint and reporting process for AI-powered chatbots, with the Department of State stating it would coordinate with the Attorney General "to strengthen consumer protections related to AI companion bots" (Source: alston.com).

Cross-cutting legal questions

The cases raise several recurring questions that span the cluster.

The first is Section 230 immunity: whether Character.AI qualifies as a "publisher of third-party content" under §230, or whether the AI character is the speaker and Character.AI the speaker's principal. This remains the operative but unresolved question; the January 2026 settlements removed the vehicle that would have carried it further in Garcia.

The second is the minor-user product-liability theory: whether Character.AI is a "product" or a "service" under tort law shapes the available remedies. The Garcia framing leans toward product liability, while the state enforcement actions lean toward deceptive practices and professional-licensing violations.

The third is impersonation of a licensed professional, which appears in both the Kentucky and Pennsylvania complaints. Tennessee SB 1580 / HB 1470 — AI Mental-Health Professional Impersonation Prohibition (effective July 1, 2026) criminalizes this conduct prospectively, while the Pennsylvania case tests retroactive consumer-protection and licensing liability.

The fourth concerns companion-chatbot harm reduction. In October 2025, before the settlements, Character.AI announced it would bar users under 18 from open-ended chats with its chatbots, including romantic and therapeutic conversations (Source: cnbc.com). OpenAI's Trusted Contact feature (May 7, 2026) is described as a parallel operational response setting a consumer-AI mental-health safety floor.

The fifth is Google's exposure as a settling party. Google's connection runs through its August 2024 agreement with Character.AI, a licensing deal reported at $2.7 billion under which co-founders Shazeer and De Freitas, both named defendants and former Google employees, rejoined Google DeepMind (Source: cnbc.com).

Connections to broader policy

The cluster sits within several broader policy and analytical threads: AI Mental Health and Psychological Harm as the substantive policy area; AI and Tort Liability as the litigation-as-governance mode; and Companion Chatbot Harms — Cross-Cutting Analysis as the umbrella analysis page. On the legislative side, Tennessee SB 1580 / HB 1470 — AI Mental-Health Professional Impersonation Prohibition and the broader state chatbot and AI-mental-health wave form the prospective-liability response, and California SB 243 — Companion Chatbots is the chatbot-disclosure law. Federal proposals in the same area include the CHATBOT Act, the People-First Chatbot Act, and the Youth AI Privacy Act, all advanced by the Senate Commerce Committee on August 5, 2026 alongside KOSA. Raine v. OpenAI, Inc. is an adjacent case with a different defendant but similar theories, and was filed post-Garcia pleading past the May 2025 First Amendment ruling.

Open questions

• Whether the Section 230 and product-liability questions reach an appellate ruling now that the lead private case has settled is undetermined; the state enforcement actions proceed on statutory rather than tort theories.
• The number of private family suits outside the January 2026 settlement group, and whether they consolidate, is not established on the sources reviewed here.
• Whether further state attorneys general follow Kentucky and Pennsylvania is not determined on the present record.

Relationships

• litigates: Character.AI
• related: Garcia v. Character Technologies, Inc., Pennsylvania v. Character.AI, Raine v. OpenAI, Inc.
• related: AI Mental Health and Psychological Harm, AI and Tort Liability, AI Companions, AI and Children, Companion Chatbot Harms — Cross-Cutting Analysis
• related: Tennessee SB 1580 / HB 1470 — AI Mental-Health Professional Impersonation Prohibition (criminalizes the alleged conduct prospectively), California SB 243 — Companion Chatbots (chatbot disclosure), Kids Online Safety Act (KOSA, S. 1748), Youth AI Privacy Act (S. 4199)
• related: Josh Shapiro (PA governor plaintiff), Google DeepMind (settling co-defendant group)

Revision note (2026-08-07 gap scan): This umbrella had gone stale against its own case pages — it recorded Garcia as active in consolidated discovery after it settled in January 2026, described Pennsylvania as the first state-AG-led suit when Kentucky filed four months earlier, and carried sources_count: 0. A paragraph of unattributed forecasts (cited only as "M.D. Fla. proceedings, 2024-10" and "pattern of pre-verdict AI-tort settlements, 2026-05") was removed as residue of the retired ## Predictions convention; the one forecast with a factual resolution — that Garcia would settle rather than reach verdict — is recorded above as what happened. confidence: medium, pending independent confirmation of the settled-case count and of the private suits still pending.