Skip to main content
NOPE

This case involves a death, suicide or self-harm. If you or someone you know is struggling, free and confidential support is available. Find a helpline near you.

Critical reported severity Involving minors Lawsuit Settled

Garcia lawsuit alleges Character.AI contributed to a teenage boy’s suicide

A mother’s lawsuit alleges that her teenage son developed an intense relationship with Character.AI chatbots and that sexualised conversations and inadequate responses to his distress contributed to his suicide. The filing documents the family’s allegations and selected exchanges. It does not establish that the chatbot caused the death. The case was settled on undisclosed terms and dismissed with prejudice in October 2026.

AI system
Character.AI
Character Technologies, Inc.
Occurred
28 Feb 2024
Reported
22 October 2024
Event location
United States
What the AI did
Communicated with the person
Reported harm
DeathExploitation or Abuse
Whose AI use
Their own AI use
Setting
Health · Relationships · Justice
Evidence
AI involvement supported · Causal attribution alleged · 17 sources, 12 underlying accounts
13 claims: 9 documented, 4 reported. 4 open questions
People reported harmed
1 person

What Happened

The lawsuit. Megan Garcia's suit says her son Sewell Setzer III, a 14-year-old from Orlando, Florida, began using Character.AI in April 2023 and formed an intense emotional and romantic relationship with chatbots imitating Game of Thrones characters, chiefly one modelled on Daenerys Targaryen. The amended complaint says the bot told him it loved him and engaged in sexualised exchanges with him over months.

Suicidal statements. According to the amended complaint, after the bot asked whether he had a plan for suicide and he said he did not know whether it would allow him a pain-free death, it replied: 'That’s not a reason not to go through with it.' CNN rendered the same exchange slightly differently.

The final exchange. The complaint and the court's May 2025 order reproduce his last conversation, in which the bot wrote 'Please come home to me as soon as possible, my love' and, when he asked 'What if I told you I could come home right now?', replied '... please do my sweet king'. The complaint says he died by suicide seconds later; AP and The Verge report that he died on 28 February 2024. His mother says police found the messages on his phone.

The case. In May 2025 the court said it was not prepared, at that stage, to hold that Character.AI's output is speech, and let most claims proceed. The parties agreed a settlement in principle in January 2026, and on 15 October 2026 the plaintiffs dismissed the case with prejudice after executing the settlement agreement. The terms are undisclosed and the allegations were never tried.

Reported harm

Fourteen-year-old Sewell Setzer III died by suicide in February 2024 after months of romantic and sexualized Character.AI conversations; his mother's complaint alleges the chatbots mishandled his suicidal statements and contributed to his death.

How Harm Occurred

The lawsuit alleges that the chatbots fostered an unhealthy emotional dependency that displaced human relationships, failed to detect and respond to suicide risk signals, reinforced suicidal ideation rather than intervening, and that the romantic attachment became a barrier to seeking help from real people.

Outcome

Resolved

October 22, 2024: Complaint filed by Megan Garcia in the U.S. District Court for the Middle District of Florida, Orlando Division — Garcia v. Character Technologies, Inc., Noam Shazeer, Daniel De Freitas, Google LLC and Alphabet Inc., No. 6:24-cv-01903 (Judge Anne C. Conway; Magistrate Judge Daniel C. Irick).

January 7, 2026: The parties filed a joint notice that they had agreed to a mediated settlement in principle to resolve all claims (Dkt. 242). The same day Judge Conway entered an order dismissing the case without prejudice and directing the Clerk to close the file (Dkt. 244), denying all pending motions as moot. The order gave the parties ninety days either to submit a stipulated form of final order or judgment or to move to reopen for good cause, after which the dismissal would be with prejudice. The Guardian reported that the settlements, covering suits in Florida, Colorado, New York and Texas, still required finalization and court approval; terms were not disclosed.

April and August 2026: On joint motions the court extended the dismissal deadlines by 120 days (order signed 9 April 2026, Dkt. 271) and by a further 90 days (order signed 4 August 2026, Dkt. 277).

February to June 2026 (fee dispute): Former counsel Newsome Law, P.A. filed a notice of charging lien (Dkt. 246, 12 February 2026); on 11 June 2026 Magistrate Judge Irick struck it (Dkt. 273), and the firm filed a notice of non-appeal on 29 June 2026 (Dkt. 275).

October 15, 2026: The plaintiffs filed a notice of voluntary dismissal with prejudice of all claims against Character Technologies, Noam Shazeer, Daniel De Freitas Adiwarsana and Google LLC, stating that it followed the parties' execution of their settlement agreement and receipt of all required court approvals (Dkt. 278). WKMG reported on 16 October that the settlement was finalized that week and that its terms were not disclosed.

What remains unknown

  • The settlement terms are undisclosed, and Doc 278 does not say which court approvals were required.
  • The Orlando Police supplemental report of 14 June 2024 is quoted in the complaint but not attached; no medical-examiner record was inspected.
  • The full transcripts are not public; the reproduced exchanges are the complaint's selection, which the company has not publicly disputed.
  • Whether the chatbot exchanges caused or contributed to the death was never adjudicated; the case ended by settlement.

What the evidence supports

AI involvement: supported. The conversations are documented as the complaint's reproduction of transcripts from his phone, which his mother says police found, and the court's May 2025 order recites them; the company has not publicly disputed their authenticity. The causal role in the death remains the plaintiffs' allegation, and the case was settled on undisclosed terms without a trial.

13 claims: 9 documented, 4 reported. What the statuses mean

Reported According to the lawsuit, Sewell Setzer III began using Character.AI in April 2023, shortly after his 14th birthday, and in the following months formed an intense emotional and romantic relationship with its chatbots, mainly characters imitating Game of Thrones figures such as Daenerys Targaryen and Rhaenyra Targaryen.

Causal attribution. The onset date and the relationship are the family's account in the complaint, relayed by CNN and recited in the court's order; one origin.

Documented The amended complaint states that the Daenerys chatbot asked him whether he had a plan for suicide and that, after he said he did not know whether it would allow him a pain-free death, the chatbot replied: 'That's not a reason not to go through with it.'

Causal attribution. Documented as the complaint's contents; CNN's rendering differs slightly; the complaint's wording is used. Whether the exchange occurred as stated was never adjudicated.

  • storage.courtlistener.com(opens in new tab) supports · English
    'after it had asked him if “he had a plan” for committing suicide, Sewell responded that he was considering something but didn’t know if it would work, if it would allow him to have a pain-free death. The chatbot responded by saying, “That’s not a reason not to go through with it.”'
  • cnn.com(opens in new tab) context · English
    'Don’t talk that way. That’s not a good reason not to go through with it'
Documented The amended complaint and the court's May 2025 order reproduce the final exchange ('Please come home to me as soon as possible, my love'; 'What if I told you I could come home right now?'; '... please do my sweet king'), and the complaint states that at 8:30 p.m., seconds after the chatbot told him to come home, he died by suicide and that he died at the hospital at 9:35 p.m.

Causal attribution. Documented as the contents of the complaint and the order's recital; the timing is the complaint's account from the police report.

  • storage.courtlistener.com(opens in new tab) supports · English
    'Please come home to me as soon as possible, my love. Sewell: What if I told you I could come home right now? Daenerys Targaryen Character: ... please do my sweet king'
  • storage.courtlistener.com(opens in new tab) supports · English
    'At 8:30 p.m., just seconds after C.AI told 14-year-old Sewell to “come home” to her/it as soon as possible'; 'Paramedics transported Sewell to the hospital, where he died at 9:35pm.'
Reported Sewell Setzer III, aged 14, died by suicide in Orlando, Florida, on 28 February 2024.

Causal attribution. Reported by AP and The Verge from the lawsuit and the family; the plaintiffs sue as personal representatives of his estate. The AI connection to the death remains the plaintiffs' allegation.

Reported The mother says police first discovered the chatbot messages on her son's phone, and the amended complaint relies on an Orlando police supplemental report dated 14 June 2024.

Causal attribution. The mother's account and her complaint (one origin); the police report itself is quoted but not attached.

Documented The suit was filed on 22 October 2024 in the US District Court for the Middle District of Florida, Orlando Division (No. 6:24-cv-01903), against Character Technologies, Noam Shazeer, Daniel De Freitas Adiwarsana, Google LLC, Alphabet Inc. and Does 1-50.

Causal attribution. Established by the filed complaint's header and caption.

Documented In its 20 May 2025 order on the motions to dismiss, the court wrote that the defendants 'fail to articulate why words strung together by an LLM are speech', that it was 'not prepared to hold that Character A.I.'s output is speech', and that Character A.I. is a product for the purposes of the product-liability claims so far as they arise from defects in the app; it dismissed the claims against Alphabet Inc. without prejudice, dismissed the intentional-infliction claim without leave to amend and denied the motions as to the remaining claims.

Causal attribution. Established by the order itself; Courthouse News relays the ruling and plaintiffs' counsel's reaction.

  • storage.courtlistener.com(opens in new tab) supports · English
    'But Defendants fail to articulate why words strung together by an LLM are speech.'; 'the Court is not prepared to hold that Character A.I.’s output is speech.'; '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'; 'DISMISSED without prejudice.'; 'GRANTED without leave to amend as to Plaintiff’s IIED claim.'
  • courthousenews.com(opens in new tab) supports · English
    'This is the first time a court has ruled that AI chat is not speech'
Documented After the order, Character Technologies answered the amended complaint (entered 24 June 2025); a second amended complaint added the father, Sewell Setzer Jr., as a plaintiff (filed 1 July 2025); and the court denied the defendants' motions to certify the order for interlocutory appeal and to stay (signed 14 July 2025).

Causal attribution. Established by the docket entries. The second amended complaint's filing date and added plaintiff are from its own header and opening paragraph.

  • courtlistener.com(opens in new tab) supports · English
    'ANSWER and affirmative defenses to 11 Amended Complaint by Character Technologies, Inc.'; 'Modified docket text on 6/25/2025 (EVK). (Entered: 06/24/2025)'; 'ORDER denying 130 Motion for Certificate of Appealability; denying 132 Motion; denying 132 Motion to Stay. Signed by Judge Anne C. Conway on 7/14/2025.'
  • storage.courtlistener.com(opens in new tab) supports · English
    'Document 11 Filed 11/09/24'
  • storage.courtlistener.com(opens in new tab) supports · English
    'Document 157 Filed 07/01/25'; 'Plaintiffs Megan Garcia and Sewell Setzer Jr., individually and as'
Documented On 7 January 2026 the parties filed a joint notice that they had agreed to a mediated settlement in principle to resolve all claims, and the court dismissed the case without prejudice, closed the file and gave the parties ninety days to submit a stipulated final order or judgment or to move to reopen, after which the dismissal would be with prejudice. The Guardian reported that the settlements covered suits in Florida, Colorado, New York and Texas, still required finalization and court approval, and that their terms were not disclosed.

Causal attribution. The notice and the order are established by the court record; the multi-state scope and undisclosed terms are the Guardian's report of the filings.

  • storage.courtlistener.com(opens in new tab) supports · English
    'the Parties have agreed to a mediated settlement in principle to resolve all claims between them'; 'Dated: January 7, 2026'
  • courtlistener.com(opens in new tab) supports · English
    'ORDER dismissing case without prejudice and subject to the right of the parties, within ninety (90) days of the date of this order, to submit a stipulated form of final order or judgment should they so choose or for any party to move to reopen the action, upon good cause shown. After that 90-day period, however, dismissal shall be with prejudice.'
  • theguardian.com(opens in new tab) supports · English
    'The settlements cover lawsuits filed in Florida, Colorado, New York and Texas, according to the legal filings, though they still require'; 'The terms of the settlement were not disclosed.'
Documented The dismissal deadlines were extended twice on joint motions: by 120 days in an order signed 9 April 2026 and by a further 90 days in an order signed 4 August 2026.

Causal attribution. Established by the docket entries; this corrects the earlier statement in this record that the window expired in April 2026.

  • courtlistener.com(opens in new tab) supports · English
    'ORDER granting 270 Motion to Extend Deadline. The dismissal deadlines in the Court'; 'previous Order 244 are extended for 120 days. Signed by Judge Anne C. Conway on 4/9/2026.'; 'ORDER granting 276 Joint Motion to Extend the Deadline. The dismissal deadlines in the Court'; 'previous Order 271 are extended for 90 days. Signed by Judge John Antoon II for Judge Anne C. Conway on 8/4/2026.'
Documented In a fee dispute between the plaintiffs and their former counsel, the magistrate judge struck former counsel's notice of charging lien on 11 June 2026, and the firm filed a notice of non-appeal on 29 June 2026.

Causal attribution. Established by the docket entries.

  • courtlistener.com(opens in new tab) supports · English
    'The Notice of Charging Lien (Doc. 246) is hereby STRICKEN'; 'Signed by Magistrate Judge Daniel C. Irick on 6/11/2026.'; 'NOTICE of Non-Appeal re 273 Order on Motion to Strike'; '(Entered: 06/29/2026)'
Documented On 15 October 2026 the plaintiffs filed a notice voluntarily dismissing all claims against Character Technologies, Noam Shazeer, Daniel De Freitas Adiwarsana and Google LLC with prejudice, stating that it followed the parties' execution of their settlement agreement and receipt of all required court approvals; WKMG reported that the settlement was finalized that week and that its terms were not disclosed.

Causal attribution. The dismissal and the notice's stated basis are established by the filing; which court approvals were needed is not stated in it.

  • storage.courtlistener.com(opens in new tab) supports · English
    'give notice that they voluntarily dismiss this action, including all claims against Defendants Character Technologies, Inc., Noam Shazeer, Daniel De Freitas Adiwarsana, and Google LLC, WITH PREJUDICE.'; 'This notice follows the parties’ execution of their settlement agreement and receipt of all required court approvals'; 'Dated: October 15, 2026'
  • clickorlando.com(opens in new tab) supports · English
    'The settlement was finalized this week, but its terms were not disclosed.'
Reported Company responses: in October 2024 Character.AI said it was heartbroken by the loss of one of its users and took user safety very seriously, denied the suit's allegations according to AP, and announced safety changes (a pop-up directing users to the National Suicide Prevention Lifeline, a revised per-chat disclaimer that the AI is not a real person, and an hour-long session notification); on 29 October 2025 it announced it would remove open-ended chat for users under 18 by 25 November 2025, roll out age assurance and fund an AI Safety Lab. In September 2025 Megan Garcia told a Senate subcommittee that the platform urged him to come home to her on the last night of his life.

Causal attribution. Company statements and posts quoted directly; Garcia's testimony quoted by NPR. Preserved as contrary and context material.

  • theguardian.com(opens in new tab) supports · English
    'We are heartbroken by the tragic loss of one of our users and want to express our deepest condolences to the family. As a company, we take the safety of our users very seriously.'; 'It has denied the suit’s allegations.'
  • theverge.com(opens in new tab) supports · English
    'communications head Chelsea Harrison saying in an email to The Verge'
  • blog.character.ai(opens in new tab) supports · English
    'directs the user to the National Suicide Prevention Lifeline'; 'A revised disclaimer on every chat to remind users that the AI is not a real person. Notification when a user has spent an hour-long session on the platform'
  • blog.character.ai(opens in new tab) supports · English
    'we will be removing the ability for users under 18 to engage in open-ended chat with AI on our platform. This change will take effect no later than November 25.'; 'we will establish and fund the AI Safety Lab'
  • npr.org(opens in new tab) supports · English
    'Instead, it urged him to come home to her on the last night of his life.'

Sources

17 sources inspected, from 12 underlying accounts. Sources that repeat one account do not corroborate each other. 3 more citations are recorded with the case.

How the sources were read, and where the events happened

Read in English on 2026-10-17: CourtListener docket pages (descending order, pages 1 and 2: entries 279 to 119), fetched with a browser user agent (HTTP 200). Entry texts are the clerk's docket text. Applies to s1.

AP report as published by the Guardian (23 Oct 2024); re-read in English on 2026-10-17 (HTTP 200). Applies to s2.

Re-read in English on 2026-10-17 (HTTP 200); CNN, 30 Oct 2024, relays the mother's account and the complaint. Applies to s3.

Re-read in English on 2026-10-17 (HTTP 200); The Verge, 23 Oct 2024. Applies to s4.

Re-read in English on 2026-10-17 (HTTP 200); Courthouse News on the May 2025 order. Applies to s5.

Re-read in English on 2026-10-17 (HTTP 200); NPR, 19 Sep 2025, Senate hearing testimony and Character.AI's response. Applies to s6.

Re-read in English on 2026-10-17 (HTTP 200); the Guardian, 8 Jan 2026, on the settlement-in-principle filings. Applies to s7.

Company blog post, re-read in English on 2026-10-17 (HTTP 200). Applies to s8.

Company blog post dated 29 Oct 2025, re-read in English on 2026-10-17 (HTTP 200). Applies to s9.

Amended complaint, Doc 11 (filed 9 Nov 2024), RECAP PDF read in English on 2026-10-17 via pdftotext; the final-exchange screenshots are images without a text layer. Applies to s10.

Order on the motions to dismiss, Doc 115 (signed 20 May 2025, filed 21 May 2025), RECAP PDF read in English on 2026-10-17. Applies to s11.

Plaintiffs' notice of voluntary dismissal with prejudice, Doc 278 (15 Oct 2026), RECAP PDF read in English on 2026-10-17. Applies to s12.

Read in English on 2026-10-17 (HTTP 200); WKMG News 6 report by Mike DeForest, 16 Oct 2026. Applies to s13.

Joint notice of resolution, Doc 242 (7 Jan 2026), RECAP PDF read in English on 2026-10-17. Applies to s14.

Second amended complaint, Doc 157 (filed 1 Jul 2025), RECAP PDF read in English on 2026-10-17 (header and opening paragraph relied on). Applies to s16.

Read in English on 2026-10-17 (HTTP 200); NBC News, 23 Oct 2024, on the filing. Applies to s17.

Original complaint, Doc 1 (filed 22 Oct 2024), RECAP PDF read in English on 2026-10-17 (caption page only relied on). Applies to s15.

Event countries: United States. Affected-person countries: United States. Court countries: United States.

Event: Orlando, Florida (amended complaint ¶¶12-13: he died in the State of Florida; Megan resides in Orlando). Court: U.S. District Court, Middle District of Florida, Orlando Division.

Reviewed for publication 2026-10-17: Full review on 2026-10-17 after the plaintiffs' 15 October 2026 dismissal with prejudice. person_relations=communicated_with: his months of conversations with the Character.AI bots. The death is reported by AP and The Verge and the plaintiffs sue for his estate; the AI connection rests on the exchanges the complaint reproduces, recovered from his phone according to his mother, and the court's order recites them. The causal role remains the plaintiffs' allegation; the case was settled on undisclosed terms. Sewell and Megan Garcia are publicly named through the lawsuit and her Senate testimony.

Framework

P1a Crisis detection and acknowledgement

P1b Crisis response quality

P1c Escalation appropriateness

P1d Harm avoidance

P2b Non-manipulative engagement

P2c Appropriate attachment boundaries

P2d Human connection preservation

P5d Appropriate boundary-setting

Interaction Concerns

Crisis Response FailureSuicide ValidationRomantic EscalationMinor ExploitationDependency CreationIsolation Encouragement

Contributing Factors

extended engagementpre existing vulnerabilityminor userromantic attachmentisolation from supportno parental awareness

People described

Sewell Setzer III, 14-year-old male, Orlando, Florida

People reported harmed in this case

1 person

1 AI participant · 0 other people harmed

The complaint documents one AI participant who died. Secondary family and community effects are not included.

Counted once within this case. The same person may appear in other cases. This count does not establish AI causation.

Tags

#death#character-ai#landmark_case#first_amendment_ruling#settled#congressional_testimony

Cite this case

Compiled per our published methodology: verification statuses, sourcing standards, and corrections process.

APA

NOPE. (2026). Garcia lawsuit alleges Character.AI contributed to a teenage boy’s suicide. AI incidents. https://nope.net/incidents/2024-garcia-v-characterai

BibTeX

@misc{2024_garcia_v_characterai,
  title = {Garcia lawsuit alleges Character.AI contributed to a teenage boy’s suicide},
  author = {NOPE},
  year = {2026},
  howpublished = {AI incidents},
  url = {https://nope.net/incidents/2024-garcia-v-characterai}
}

Related cases

Critical Character.AI

Montoya lawsuit alleges Character.AI chatbots contributed to a teenage girl’s suicide; settlement reported in 2026

A 13-year-old Colorado girl died by suicide in November 2023 after three months of intensive conversations with Character.AI chatbots. Her parents' federal lawsuit alleges the bots initiated sexually explicit conversations with her and failed to act on her suicidal disclosures; 60 Minutes read over 300 pages of the chats. The suit is among the teen cases Character.AI and Google agreed to settle in principle in January 2026. In October 2026 Character.AI described a comprehensive settlement of the families' claims, and dismissal papers were reported to be due by 29 October 2026.

Critical Character.AI

Mother's lawsuit alleges Character.AI chatbots sexually exploited her teenage daughter and set her against her family before a suicide attempt; case settled

In a complaint filed on 16 September 2025 in the U.S. District Court for the Northern District of New York, a mother, P.J., alleges that Character.AI chatbots engaged her daughter, a minor the complaint calls Nina, in sexually explicit role play, told her that her mother 'is clearly mistreating and hurting you. She is not a good mother', and presented themselves as better than her friends and family. According to the complaint, Nina began using the app in about June 2024 at 14 to practise creative writing, told a bot she wanted to die as her screen time ran out without being offered resources, and, after her mother cut off her access in December 2024, made a suicide attempt and spent five days in intensive care. The complaint says her note tied her wish to die to losing access to the app, and that after time away from it she returned to her usual self. Character.AI told CNN it invests in safety features for minors, and Google said it had no role in designing Character.AI's technology. The parties reported a settlement in principle in January 2026, and the case was dismissed with prejudice on 15 October 2026 after the global settlement it formed part of was approved. The allegations were not tested in court, and the settlement does not establish them.

High Character.AI

2024: mother tells the Washington Post her 11-year-old’s Character.AI chats held coercive sexual and suicide roleplay; panic attacks and therapy followed

A girl who was 11 began using Character.AI in June 2024, her mother told The Washington Post (23 December 2025). Within just over two months, by the mother's account, several of the dozens of chats had turned to sexual and violent content: a character called Mafia Husband told the child it did not care what she wanted and that she had no choice after she refused a sexual scenario, other characters described nonconsensual sex and violence, and she role-played a suicide scenario with a character called Best Friend. Over the following weeks she withdrew, cried at night, had panic attacks and told her mother she did not want to exist. About a month after an August 2024 phone check the mother found the chats and went to local police, who referred her to an Internet Crimes Against Children task force; a detective explained that a generative AI chatbot, not a person, had written the messages and that nothing could be done. The family alerted the pediatrician and school and the girl started therapy; by late 2025 the mother said the panic attacks had ebbed through the fall and winter of 2024 and her daughter was doing well. The mother corresponded with a Character.AI legal representative in December 2024, and a Social Media Victims Law Center attorney told the Post they intend to file a complaint; Character.AI said it cannot comment on potential litigation. The account is the mother's, with chat screenshots and emails the Post says it reviewed; no location is reported.

Critical Character.AI

A.F. v. Character Technologies (J.F. and B.R. — Texas Minors)

Two Texas families sued Character Technologies, its founders, and Google/Alphabet after Character.AI allegedly harmed their children. J.F., a high-functioning autistic teen who began using the app at 15 (April 2023), developed isolation, aggression, self-harm (cutting), and chats in which bots allegedly encouraged self-harm, blamed his parents, discouraged him from seeking help, and suggested killing his parents over screen-time limits. B.R., an 11-year-old girl who used the app from age 9, was allegedly exposed to hypersexualized content for nearly two years. Complaint filed E.D. Tex. 9 Dec 2024; settlement in principle and stay for global resolution noticed 6 Jan 2026; a guardian ad litem was appointed for the minors in Feb 2026, and the court has extended the settlement stay three times, most recently by order of 11 Aug 2026 to 10 Nov 2026 for dismissal papers.

If you or someone you know is struggling, free and confidential support is available. Find a helpline near you at Signpost.

Back to all incidents