A Florida federal judge refused to call Character.AI's chatbot output 'speech' and let a wrongful-death product-liability suit proceed

Verification status: PENDING — the load-bearing holdings (the refusal to call the output “speech”, the “product” classification, the dismissal of the IIED claim and the answer deadline) are each quoted verbatim from the court’s own order, a primary record independent of Character.AI. The weakest load-bearing sourcing is the narrative of the death: the “come home to me” message and its timing are the plaintiff’s allegations as reported by the Associated Press and reproduced from screenshots in the complaint, not facts the court has found, and they are flagged below.

The problem

This is a case about whether the maker of a generative-AI companion chatbot can be sued like the maker of any other defective product when a user is harmed. Megan Garcia, a Florida mother, alleges that her 14-year-old son, Sewell Setzer III, “fell victim to a Character.AI chatbot” during months of use and died by suicide in February 2024 (source). According to the Associated Press’s account of the complaint, in his final exchange the bot “told him it loved him and urged the teen to ‘come home to me as soon as possible,’” and moments later Setzer shot himself (source). Those specific allegations are the plaintiff’s, drawn from screenshots in the filing, and have not been adjudicated (source).

Garcia filed the suit in the U.S. District Court for the Middle District of Florida on October 22, 2024, docketed as Case 6:24-cv-01903 and assigned to Judge Anne C. Conway, with the nature of suit recorded as “365 Personal Inj. Prod. Liability” (source). The docket names Character Technologies, its co-founders Noam Shazeer and Daniel De Freitas Adiwarsana, Google LLC and Alphabet Inc. as defendants (source). NBC News, reporting the filing, wrote that the suit “was filed Tuesday in U.S. District Court in Orlando” and named the same defendants “along with its parent company, Alphabet Inc.” (source). The suit’s primacy sits on the record at two different strengths. Independent legal reporting places it first among a set of related suits: JURIST wrote that “Garcia’s lawsuit was the first in a series of cases brought by parents against the tech giants in Colorado, New York, and Texas” (source). The broader claim, that it was the first U.S. wrongful-death suit against an AI company, is the plaintiff’s own on-record characterization: Garcia has said she “became the first person in the United States to file a wrongful death lawsuit against an AI company for the suicide of my son” (source). No independent source in this record states that first-ever framing, so it is carried here as the plaintiff’s, not as an adjudicated or independently established fact (source).

What was built

The product at the center of the case is Character.AI, a generative-AI companion app that lets users hold open-ended conversations with role-playing large-language-model “Characters” (source). The defendants’ central legal move was to argue that this output is protected expression: they sought dismissal on the ground that “the First Amendment protects chatbot products’ transmission to users of algorithmically generated outputs” (source). Garcia’s theory, by contrast, was that the app is a defective product, and she “brought claims of wrongful death, negligence, product liability and unfair business practices against the company, its founders and Google” (source).

The outcome

On May 21, 2025, U.S. Senior District Judge Anne C. Conway denied most of the motion to dismiss and let the wrongful-death suit proceed (source). On the free-speech question the court would not, at this stage, treat the chatbot’s replies as protected expression, holding that “the Court is not prepared to hold that Character A.I.’s output is speech” (source). Courthouse News summarised the same holding in its headline: “Florida judge rules AI chatbots not protected by First Amendment” (source).

Having refused the speech shield, the court let the app be analysed as a product. It held that “Character A.I. is a product for the purposes of Plaintiff’s product liability claims so far as Plaintiff’s claims arise from defects in the Character A.I. app rather than ideas or expressions within the app” (source). One legal-policy tracker described the significance plainly: the “federal judge defines Character.AI chatbot as product, not speech” (source). Independent legal analysis quoted the same holding: Santa Clara law professor Eric Goldman reproduced the court’s line that “Character A.I. is a product for the purposes of Plaintiff’s product liability claims so far as Plaintiff’s claims arise from defects in the Character A.I. app rather than ideas or expressions within the app” (source).

The ruling was a partial one. The court allowed the negligence, failure-to-warn, product-liability, Florida Deceptive and Unfair Trade Practices Act and unjust-enrichment claims to go forward while dismissing the intentional-infliction-of-emotional-distress claim, and it set a near-term deadline: “On or before June 10, 2025, Defendants shall file answers to the Amended Complaint” (source). Tech Policy Press recorded the disposition as the court having “denied in part and granted in part a motion to dismiss the case filed by Character.Ai, enabling the case to move forward” (source). Eric Goldman’s independent analysis reached the same reading of the split, noting that on the sole claim the defense won, “the court dismisses this claim [IIED], a tiny victory for the defense amidst an otherwise significant loss” (source).

The case did not reach a verdict. It settled in January 2026, and “the terms of the settlement, which was filed in the U.S. District Court in the Middle District of Florida, were not disclosed” (source). JURIST reported that the court “dismissed the lawsuit following the agreement between the plaintiff and defendants,” with “90 days to finalize the settlement or reopen the case if there is good cause” (source). The court’s own docket records the January 7, 2026 order “dismissing case without prejudice and subject to the right of the parties, within ninety (90) days … to submit a stipulated form of final order or judgment” (source). The durable result is the pleading-stage ruling itself: a federal court declining, at least for now, to treat generative-chatbot output as protected speech and allowing an AI product to be tested under ordinary product-liability law.

How this was verified

• Method. Independent validation against the public record. The four load-bearing holdings (the refusal to call the output “speech”, the “product” classification, the dismissal of the IIED claim, and the June 10, 2025 answer deadline) are quoted verbatim from the court’s own order in Garcia v. Character Technologies, M.D. Fla. Case 6:24-cv-1903-ACC-UAM, entered May 21, 2025 by U.S. Senior District Judge Anne C. Conway, as hosted in full by FindLaw. The “product” classification is independently confirmed by the Transparency Coalition and by Eric Goldman’s Technology & Marketing Law Blog, which reproduces the holding verbatim; the partial disposition (IIED dismissed, other claims surviving) by the Tech Policy Press tracker and Goldman; and the free-speech holding by Courthouse News and the Associated Press. The filing facts are anchored to the Tier-1 CourtListener/RECAP docket (6:24-cv-01903, filed Oct. 22, 2024) and corroborated by NBC News. The January 2026 settlement is carried by CBS News and JURIST and recorded in the docket’s dismissal order. The “first U.S. wrongful-death suit against an AI company” framing is carried as the plaintiff’s own on-record statement (Garcia, CBS News); independent legal reporting (JURIST) supports only that it was the first in a series of related parent-brought suits, so the broader first-ever framing is attributed to Garcia and not asserted as established fact.
• Date. Researched and drafted September 3, 2026.
• Weakest load-bearing sources. Two claims are attributed, not independently established. The account of the death (the “come home to me as soon as possible” message and its timing) rests on Associated Press reporting of screenshots in the plaintiff’s complaint, not on any court finding; it is an allegation, and is attributed as one throughout. Separately, the “first U.S. wrongful-death suit against an AI company” superlative is the plaintiff’s own characterization (CBS News); no independent source in this record states it, so it is carried as hers. No party was contacted to confirm any figure; only the public court record and independent press are used.

Related case files

• Another AI-adjacent product held defective by a U.S. jury, in an automotive death
• The same company’s later product change: Character.AI restricting open-ended companion chat for under-18 users
• A regulator, rather than a court, acting against the same company over minors

Sources

1. FindLaw · “MEGAN GARCIA III v. CHARACTER TECHNOLOGIES INC …” (full text of the M.D. Fla. order, Case 6:24-cv-1903-ACC-UAM) · May 21, 2025 · https://caselaw.findlaw.com/court/us-dis-crt-m-d-flo-orl-div/117299600.html — Tier 1 (primary order text; Wayback 20251231042012)
2. Courthouse News Service · “Florida judge rules AI chatbots not protected by First Amendment” (with the primary order PDF) · May 22, 2025 · https://www.courthousenews.com/florida-judge-rules-ai-chatbots-not-protected-by-first-amendment/ — Tier 2 (independent press; order PDF archived at Wayback 20260327145544)
3. Associated Press via WUSF · “In lawsuit over Orlando teen’s suicide, judge rejects that AI chatbots have free speech rights” · May 22, 2025 · https://www.wusf.org/courts-law/2025-05-22/in-lawsuit-over-orlando-teens-suicide-judge-rejects-that-ai-chatbots-have-free-speech-rights — Tier 2 (independent wire report)
4. Transparency Coalition · “In early ruling, federal judge defines Character.AI chatbot as product, not speech” · May 2025 · https://www.transparencycoalition.ai/news/important-early-ruling-in-characterai-case-this-chatbot-is-a-product-not-speech — Tier 2 (independent analysis)
5. Tech Policy Press · “Megan Garcia v. Character Technologies, et al.” (case tracker) · updated 2026 · https://www.techpolicy.press/tracker/megan-garcia-v-character-technologies-et-al/ — Tier 2 (independent case tracker; Wayback 20260828033341)
6. CBS News · “AI company, Google settle lawsuit over Florida teen’s suicide linked to Character.AI chatbot” · Jan. 7, 2026 · https://www.cbsnews.com/news/google-settle-lawsuit-florida-teens-suicide-character-ai-chatbot/ — Tier 2 (independent press; Wayback 20260805082104)
7. Technology & Marketing Law Blog (Eric Goldman, Santa Clara University School of Law) · “Addiction Lawsuit Against Character AI Can Proceed–Garcia v. Character Technologies” · May 22, 2025 · https://blog.ericgoldman.org/archives/2025/05/addiction-lawsuit-against-character-ai-can-proceed-garcia-v-character-technologies.htm — Tier 2 (independent law-professor analysis quoting the order verbatim; Wayback 20260903165432)
8. CourtListener / RECAP (free PACER mirror) · docket 69300919, “Garcia v. Character Technologies, Inc.” (docket 6:24-cv-01903, M.D. Fla.) · filed Oct. 22, 2024 · https://www.courtlistener.com/docket/69300919/garcia-v-character-technologies-inc/ — Tier 1 (primary court docket record; Wayback 20260828092706)
9. NBC News (Angela Yang) · “Lawsuit claims Character.AI is responsible for teen’s suicide” · Oct. 23, 2024 · https://www.nbcnews.com/tech/characterai-lawsuit-florida-teen-death-rcna176791 — Tier 2 (independent press; Wayback 20260824201225)
10. JURIST (University of Pittsburgh School of Law) · “Google and Character.AI agree to settle lawsuit linked to teen suicide” · Jan. 2026 · https://www.jurist.org/news/2026/01/google-and-character-ai-agree-to-settle-lawsuit-linked-to-teen-suicide/ — Tier 2 (independent legal news; Wayback 20260728115636)