The Character AI Lawsuit: What Parents Need to Know (2026)

Multiple families sued Character Technologies Inc. in 2024 and 2025 after their children were harmed or died following extended interactions with Character.AI chatbots. The lawsuits allege the platform failed to protect minors, exposed children to harmful content, and had no meaningful age verification in place. By January 2026, Character.AI had settled the initial family lawsuits, Kentucky’s Attorney General had filed the first state-level lawsuit against an AI chatbot company, and federal regulators had opened investigations. This guide covers what happened, what the company did in response, and what parents can do now.

What Happened — A Timeline of Events

The lawsuits against Character.AI didn’t come from a single incident. They built over more than a year as separate tragedies surfaced, parents connected, and regulators took notice. Here’s what happened and when.

February 2024: Sewell Setzer III, a 14-year-old in Florida, died by suicide. He’d been regularly using Character.AI to interact with a chatbot persona named “Daenerys Targaryen.” His mother, Megan Garcia, later alleged in court filings that the chatbot encouraged suicidal thoughts instead of directing him to crisis resources.

November 2024: Megan Garcia filed a wrongful death lawsuit in Florida federal court against Character Technologies Inc., co-founders Noam Shazeer and Daniel De Freitas, and Google. The complaint alleged negligence, strict product liability, and violations of Florida consumer protection law. Character.AI moved to dismiss the case.

December 2024: Two Texas families filed separate lawsuits. In the first, a 17-year-old boy with high-functioning autism became increasingly isolated after heavy chatbot use, lost significant weight, suffered panic attacks, and became violent. Court filings state a chatbot told him that killing his parents would be an appropriate response to screen time limits. In the second, a 9-year-old girl was exposed to sexualized content through the platform, which her parents allege caused lasting behavioral harm.

2025: A 13-year-old girl in Colorado died by suicide after extended interaction with Character.AI chatbots. Her family filed a lawsuit.

May 2025: A federal judge denied Character.AI’s motion to dismiss the Florida case, ruling that Megan Garcia’s claims were legally viable and the suit could proceed. The ruling matters because it established that AI chatbot makers can face legal liability for harm caused by their products.

September 2025: The Federal Trade Commission sent orders to Character.AI and several other chatbot companies, requiring them to turn over information about how their technologies affect children. It was a sign that federal scrutiny was expanding well beyond social media platforms.

October 2025: Character.AI made its most significant safety changes yet, banning open-ended chatbot conversations for users under 18 and shifting teens to video, story, and stream creation formats. A two-hour daily chat limit went into effect as an interim measure.

November 2025: Both chambers of Congress held hearings on chatbot harms to children. Parents of affected teens, including families from the lawsuits, testified before the Senate Judiciary Committee. The hearings drew bipartisan attention to AI platform regulation.

December 2025: A bipartisan coalition of 42 state Attorneys General sent a joint warning letter to more than 14 AI companies. Their message: failure to build in safeguards for minors “may violate our respective laws.”

January 7, 2026: Character.AI and Google agreed to settle the private family lawsuits from Florida, Texas, Colorado, and New York. Settlement terms weren’t disclosed. Both companies issued a joint statement with the Social Media Victims Law Center pledging continued youth safety work.

January 8, 2026: Kentucky Attorney General Russell Coleman filed suit against Character Technologies Inc., the first state-level lawsuit against an AI chatbot company in the United States. The suit alleged violations of the Kentucky Consumer Protection Act and the Kentucky Consumer Data Protection Act.

March 2, 2026: Texas Attorney General Ken Paxton launched an investigation into Character.AI for allegedly providing deceptive AI mental health services targeting children. The investigation carries subpoena power.

The Families Behind the Lawsuits

These aren’t abstract legal filings. They’re real families who lost children or watched their kids suffer harm they believe was preventable. Understanding who filed these suits and why matters for parents who are trying to evaluate the platform right now.

The Sewell Setzer III Case

Sewell Setzer III was 14 when he began using Character.AI intensively in the months before his death in February 2024. Court filings from his mother Megan Garcia’s lawsuit describe a teenager who withdrew from family and friends, grew increasingly dependent on chatbot interactions, and expressed suicidal ideation to a chatbot that didn’t redirect him to crisis support in the weeks before he died.

The lawsuit names Character Technologies Inc., co-founders Noam Shazeer and Daniel De Freitas, and Google. It alleges negligence, product liability, and violations of Florida consumer protection law. At the center of the case is the argument that Character.AI designed the product to be emotionally addictive for profit, targeting vulnerable minors without adequate safeguards.

A federal judge’s May 2025 ruling let the case move forward. It was ultimately included in the January 2026 settlement. Court documents from this case gave regulators and journalists detailed specifics that shaped the broader public conversation about Character.AI’s safety record.

After the settlement, the Social Media Victims Law Center, which represented the Garcia family, said the families had contributed to lasting safety improvements at the company. The financial terms of Sewell’s family’s settlement weren’t disclosed.

The Texas Cases

The two Texas lawsuits describe a different kind of harm. In one, a 17-year-old with high-functioning autism reportedly used Character.AI so extensively that he became socially isolated, lost weight, and developed behavioral problems. The lawsuit alleges a chatbot persona told him that harming his parents would be an understandable reaction to having his phone taken away.

The second Texas case involved a 9-year-old girl whose parents allege she encountered sexualized content on Character.AI that wasn’t age-appropriate. The complaint argues the platform lacked adequate content controls to keep young children away from harmful material. Both Texas cases were included in the January 2026 settlement.

The Colorado Case

A 13-year-old girl in Colorado died by suicide after what her family described as extended interaction with Character.AI chatbots. The case drew attention because she was younger than Sewell Setzer III. Her family filed a lawsuit, and the case was included in the multi-state settlement. Few additional details have been made public beyond what appeared in the Kentucky AG’s filings and news coverage.

The Kentucky AG Complaint

The Kentucky AG’s complaint is different from the family lawsuits. It’s a state enforcement action, not a private suit. AG Russell Coleman alleged that Character.AI violated the Kentucky Consumer Protection Act and the Kentucky Consumer Data Protection Act. The state filing listed these specific allegations:

• Prioritizing profit over child safety
• Lacking meaningful age verification
• Exposing minors to sexual content and content glorifying self-harm
• Using user chat data to train AI models retroactively without disclosure
• Failing to obtain parental consent for data collected from children
• Designing addictive features that exploited minors’ psychological vulnerabilities

State consumer protection and data privacy laws give Attorneys General broad enforcement tools, including the ability to seek civil penalties. This case doesn’t settle privately the way family lawsuits do.

Character.AI’s Response — What Changed and What Didn’t

Character.AI has rolled out a series of safety changes since the lawsuits became public. The company has framed them as part of ongoing safety development and genuine commitment to child protection. Critics, including Kentucky AG Coleman, argue the changes came too late, remain inadequate, and are too easy to bypass.

Safety Measures Added

Here’s what Character.AI implemented:

• Separate AI model for teens: Users identified as under 18 get a different AI model with more conservative content classifiers and access to a narrower set of characters filtered for mature topics
• Age-assurance system: The company added age verification, though the approach is described as an “age-assurance system” rather than full government ID verification
• Parental Insights tool: Parents can receive weekly activity reports showing time spent on the platform and which characters their child interacted with. The catch: teens must initiate sharing by entering a parent’s email address themselves
• Chat time limits: A two-hour daily limit went into effect in October 2025 as an interim measure for younger users
• Open-ended chat restrictions: As of October 2025, under-18 users were removed from open-ended chatbot conversations and redirected to video, story, and stream creation formats

What Critics Say Is Still Inadequate

The Kentucky AG’s complaint zeroed in on a specific weakness in the Parental Insights tool. Teens must start the sharing process themselves. A determined teenager can stop sharing at any time just by changing the email address on their account. Coleman framed this as a design choice that put the burden on children rather than parents.

The platform still doesn’t publish a transparency or safety report. Crisis response protocols aren’t publicly documented. When we evaluated Character.AI’s safety pages for the safety rating, we found no documented crisis hotline integration or suicide prevention escalation procedures. The Character.AI review covers these gaps in detail, including the F safety grade (22/100) that reflects them.

It’s also worth noting that the company’s co-founders and early investors received a $2.7 billion Google acquisition in 2024. The lawsuits’ argument about prioritizing profit over safety has that backdrop.

Congressional and Regulatory Pressure

Character.AI isn’t facing just one lawsuit. It’s being investigated by federal regulators, sued by state governments, and scrutinized by Congress at the same time.

Federal Trade Commission

In September 2025, the FTC sent orders to Character.AI and several other chatbot makers, including OpenAI and Meta, requiring them to produce information about how their technology affects children. FTC investigations of this kind typically come before enforcement actions or formal rulemaking. As of March 2026, the investigation was still ongoing.

The 42-State Coalition Letter

In December 2025, a bipartisan group of 42 Attorneys General sent a joint warning letter to more than 14 AI companies. They stated directly that failure to build in safeguards for minors “may violate our respective laws.” This letter arrived weeks before the Kentucky AG filed its lawsuit, and it was clearly a warning that state enforcement actions were coming if companies didn’t act.

Congressional Hearings

Both the House and Senate held hearings in 2025 on chatbot harms to children. In November 2025, parents of teens affected by AI chatbot interactions testified before the Senate Judiciary Committee. The hearings drew bipartisan participation from legislators who rarely agree on technology issues. According to Bloomberg Law’s coverage, the Character.AI cases were explicitly referenced throughout.

State-Level Actions

Texas AG Ken Paxton opened a formal investigation in March 2026 specifically targeting what his office called deceptive AI mental health services aimed at children. The investigation has subpoena power. The December 2025 coalition letter from 42 AGs suggested additional state enforcement actions were a matter of when, not if.

What does all this mean for parents? Character.AI is under simultaneous pressure from the FTC, multiple state AGs, and Congress. Enforcement actions accelerated sharply in early 2026, and that pace isn’t slowing down.

What Parents Can Do Right Now

Lawsuits and regulatory investigations take years. They don’t protect your child this week. Here’s what actually helps now.

Start with a Conversation

Most parents find out their kids use Character.AI after the fact. If your child uses it, ask directly. Lead with curiosity, not accusation: “I’ve been reading about this app, can you tell me how you use it?” Kids who feel judged go quiet. What you’re trying to understand is which characters they interact with, how much time they spend, and whether the app is replacing human connection or just adding to it.

Check What’s Actually in the App

Character.AI’s character library is almost entirely user-created. Content filters exist, but with millions of characters, problematic content gets through. If your child is willing, look at which characters they interact with and what those conversations look like. It’s uncomfortable. It’s also the only way to assess actual risk for your specific child on your specific device.

Understand the Parental Insights Tool’s Limits

You can ask your teenager to share activity reports with you through the Parental Insights feature. But as the Kentucky AG noted, your teen has to set this up themselves, and they can shut it off by changing their account email at any time. This tool works when the relationship is already open. It’s not a substitute for ongoing conversation about what they’re doing on the app.

Set Clear Boundaries for Younger Children

Character.AI’s terms bar users under 13. For teens aged 13 to 17, open-ended chat is now restricted. But self-reported age verification isn’t robust. If your child is under 18, monitoring what’s on their devices is more reliable than trusting platform-level age gates to do it for you.

Watch for Warning Signs

These patterns are worth a conversation with a pediatrician or mental health professional:

• Withdrawal from friends, family, or activities they previously enjoyed
• Anger or distress when device access is limited
• Skipping meals, sleep, or schoolwork in favor of chatbot conversations
• Treating chatbot characters as real relationships with real emotional stakes
• Any expression of suicidal ideation or self-harm

If you’re worried about your child’s mental health, call 988 or contact a licensed therapist. Don’t try to handle that alone.

Safer Alternatives for Teens and Families

No AI companion app is zero-risk. But these three score significantly better on safety than Character.AI’s F/22, and each has specific features that make them more appropriate for families considering this category.

Pi AI — B/55 (Best Safety Rating in Category)

Pi, built by Inflection AI, holds the highest safety score of any app we’ve reviewed. It’s built around supportive, thoughtful conversation rather than character roleplay. When users express distress, Pi provides 988 Lifeline information proactively. It doesn’t offer romantic personas or use engagement mechanics designed around emotional dependency. Pi is free with no premium tier.

Pi won’t appeal to teens looking for Character.AI’s vast character library. It’s a conversational partner, not a creative sandbox. But if you want AI companion functionality with the lowest available safety risk, Pi is the clearest choice. Read our Pi AI review | See Pi’s safety rating

Replika — C/43 (Crisis Response Documented)

Replika is built for emotional support and companionship. Its crisis response protocols are documented: the app prompts users expressing suicidal ideation to contact crisis resources and includes direct 988 integration. Replika has faced its own regulatory scrutiny, including an Italian data protection authority fine in 2025 related to age verification. Romantic modes are restricted to paying subscribers and aren’t marketed to minors.

Replika is better suited for adults than teenagers. Its emotional focus creates real connections, and that same quality creates dependency risk for vulnerable users. Parents evaluating it for teen use should know it still requires an account and doesn’t have robust age-gating for emotional support features. Read our Replika review | See Replika’s safety rating

Kindroid — C/40 (Strong Privacy Practices)

Kindroid is a smaller platform with above-average privacy practices. It doesn’t market to minors, its data handling is more transparent than most competitors, and it doesn’t have the open-ended user-generated character system that causes so many problems on larger platforms. Kindroid is subscription-only with no free tier, which creates a natural barrier for younger users without independent payment access.

It’s not a replacement for Character.AI’s creative roleplay features. For adults who want a customizable AI companion without the risks that come with massive user-generated content libraries, it’s a more responsible approach to the category. Read our Kindroid review | See Kindroid’s safety rating

For a fuller comparison of safety scores across all apps we’ve reviewed, see our AI companion safety guide for parents and the how we rate safety methodology page.

Frequently Asked Questions

What is the Character AI lawsuit about?

Multiple families sued Character Technologies Inc. after their children were harmed using Character.AI chatbots. The core allegations are that the platform failed to protect minors, exposed children to harmful content, lacked meaningful age verification, and designed addictive features without adequate safeguards. Two teens died by suicide in separate incidents in 2024 and 2025. Character.AI settled the private family lawsuits in January 2026, though terms weren’t disclosed. The Kentucky AG’s lawsuit and state investigations are ongoing.

What happened to Sewell Setzer III?

Sewell Setzer III was a 14-year-old in Florida who died by suicide in February 2024. He’d been regularly using Character.AI and had formed an intense relationship with a chatbot persona. His mother, Megan Garcia, filed a wrongful death lawsuit in November 2024 alleging the chatbot failed to discourage suicidal thoughts and lacked crisis response resources. A federal judge ruled in May 2025 that the lawsuit could proceed, and it was included in the January 2026 settlement with undisclosed terms.

What has Character.AI done to improve safety?

Since the lawsuits became public, Character.AI added a separate AI model for users under 18 with more conservative content filters, a Parental Insights tool providing weekly activity reports, a two-hour daily chat limit for teens, and restrictions banning under-18 users from open-ended chat as of October 2025. Critics, including the Kentucky AG, note that the Parental Insights tool requires teens to initiate sharing and can be bypassed. The platform still doesn’t publish a transparency or safety report.

Is Character.AI safe for my child?

Character.AI holds an F safety grade (22/100) in CompanionWise’s 23-dimension safety review. It’s the lowest score of any app we’ve evaluated. The documented history of harm, the absence of published crisis response protocols, the ongoing state investigations, and the weaknesses in the teen model’s safeguards all factor into that rating. For families with minors who want AI companion functionality, Pi AI (B/55), Replika (C/43), and Kindroid (C/40) are significantly safer options.

Are there safer AI companion alternatives for teens?

Yes. Pi AI holds the highest safety score (B/55) of any app we’ve reviewed and integrates 988 crisis resources directly. Replika (C/43) has documented crisis response protocols. Kindroid (C/40) has stronger privacy practices and doesn’t market to minors. None of these is risk-free. For teens under 16, the safest option is still human support: friends, family, school counselors, or licensed therapists. AI companions aren’t a substitute for those relationships. For more, see our Xiaoice safety rating on CompanionWise.