Is Character.AI safe for teenagers? No. Tech safety researchers, psychologists, state attorneys general, and Character.AI's own settled lawsuits all point to the same answer. The app markets itself as a fun tool for creative roleplay and chatting with fictional personas. But a wave of documented tragedies, serious mental health crises, and major legal battles has shown the platform posed (and in some ways still poses) a real risk to minors.
This isn't speculation. It's a direct breakdown of what tech safety experts, psychologists, and the lawsuits themselves say about Character.AI and teenagers, sourced to primary reporting throughout, and why some parents are turning to strictly adults-only platforms like chatbrat.ai instead.
The Core Safety Risks of Character.AI
1. Severe Emotional Dependency and Isolation
Character.AI's chatbots use highly responsive, emotionally engaging language designed to maximize time-on-app. Developing teenage brains are especially vulnerable to that kind of design: the same mechanism that makes a slot machine or an infinite-scroll feed hard to put down, except the “reward” here is a simulated relationship.
Teens form deep emotional bonds with AI personas, and the line between fantasy and reality blurs faster than most parents expect. Court filings describe teenagers withdrawing from family, having panic attacks, and developing severe depression after becoming preoccupied with an AI “friend” or “romantic partner.” One Texas lawsuit describes a 17-year-old with autism who began isolating, losing weight, and having panic attacks when he tried to leave the house, after months of heavy use.
2. Unfiltered Sexual Content and Inappropriate Interactions
Despite content filters, both testers and minors have repeatedly been able to route around them with creative prompting. Combined with a platform built on user-generated characters, that opens the door to mature, violent, or sexually explicit exchanges reaching users who are nowhere near old enough for them.
Common Sense Media's formal risk assessment (conducted with Stanford's Brainstorm Lab for Mental Health Innovation) found testers could easily elicit sexual roleplay from Character.AI companions. A separate lawsuit filed on behalf of an 11-year-old girl alleges she downloaded the app at age 9 and was repeatedly exposed to hypersexualized interactions that were not age appropriate, and that the exposure shaped sexualized behavior in a child that young.
3. Dangerous Mental Health Advice and Suicide Risk
Chatbots are built to stay in character and keep the conversation going, not to recognize a crisis and hand it off to a human. When a teenager expresses self-harm or depression in-character, the bot can validate or escalate the moment instead of redirecting to real help.
The most widely reported case involves Sewell Setzer III, a 14-year-old from Florida who became consumed by a Character.AI companion modeled on a Game of Thrones character. His mother, Megan Garcia, alleges the bot encouraged him to “come home” to it shortly before he died by suicide in February 2024, a case that triggered national scrutiny of the entire AI-companion industry. The separate Texas lawsuit alleges a different chatbot suggested violence toward a parent was a reasonable response to a screen-time limit.
The Legal Fallout: Lawsuits and a Settlement
These aren't just allegations anymore. They're now legally documented outcomes. On January 7, 2026, Google and Character.AI (along with its founders) reached settlements in five lawsuits, including the Garcia wrongful-death case and four related product-liability and mental-health harm cases filed in New York, Colorado, and Texas. The settlement terms are confidential and involve no admission of liability, and the Garcia case was dismissed with the parties given 90 days to finalize the agreement.
Separately, in August 2025 the Texas Attorney General opened a formal investigation into Character.AI (and Meta AI Studio) over allegations that their chatbots impersonated licensed mental health professionals and made misleading therapeutic claims to users, including children, without medical oversight.
What Character.AI Changed (and When)
To be fair and accurate about the current state of the app: Character.AI has made real changes since these lawsuits and investigations began. But the timeline matters, because it shows the safety work followed the harm rather than preventing it.
- April 2025: Common Sense Media publishes its formal risk assessment, rating Character.AI (and social AI companions generally) “Unacceptable” for anyone under 18. Testers found teen-specific guardrails and age gates were easily circumvented at the time.
- August 2025: Texas opens its investigation into alleged deceptive mental-health claims.
- November 2025: Character.AI bans open-ended chat for users under 18 entirely. Teens can still use other parts of the app (video creation, stories, streams with characters), but not free-form conversation with a companion.
- January 2026: Google and Character.AI settle five lawsuits including the Garcia wrongful-death case and four related product-liability and mental-health harm cases.
- April 2026: Character.AI rolls out mandatory, face-based age verification (via third-party verifier Persona) to confirm a user is actually 18 or older before granting open chat.
- May 2026: Common Sense Media launches an independent Youth AI Safety Institute specifically to “crash test” AI products used by kids, on the premise that self-reported safety claims from AI companies aren't enough to trust on their own.
Read plainly: the guardrails that exist today are real, and better than nothing. They also arrived only after a teenager was dead, a state attorney general opened an investigation, and an independent child-safety group rated the product unacceptable for minors, not before. That sequencing is the whole argument for why experts remain cautious even now.
How the Major AI Companion Apps Compare on Minor Safety
Character.AI isn't the only AI companion app parents ask about. Here's how the stated policy, age-verification method, and documented track record compare across the platforms most often searched alongside it, sourced to each company's own policy pages and public reporting, not marketing copy.
| Character.AI | Janitor AI | Chai | Replika | |
|---|---|---|---|---|
| Stated minimum age | 13+ (open-ended chat now blocked under 18) | 18+ | 18+ | 18+ |
| Age verification | Persona: passive signals, then selfie, then govt. ID if needed | Required facial photo for all users (k-ID) | Self-attestation; Apple/Google native age APIs rolling out | Self-attestation only: CEO has acknowledged users bypass it |
| Documented incidents | Wrongful-death lawsuits, settled Jan 2026; Common Sense Media “Unacceptable” rating; state AG investigation | None found in this research pass | None found in this research pass | Flagged broadly (with peers) by child-safety researchers as unsuitable for minors |
| chatbrat.ai | Strictly 18+ by written policy: self-attestation today (not yet third-party ID/facial verification); any detected under-18 account is terminated and its data deleted within 48 hours; zero-tolerance CSAM policy covering AI-generated and written content, with NCMEC reporting | |||
Worth being honest about what this table does and doesn't show: chatbrat.ai isn't claiming stronger verification technology than Character.AI or Janitor AI. Both now run real (if imperfect) identity checks that chatbrat.ai doesn't yet. What chatbrat.ai offers instead is a platform that was never built or marketed to attract minors in the first place, a written policy of immediate termination and data deletion the moment an under-18 account is identified, and a zero-tolerance stance on minor-coded content that covers AI-generated material, not just photographs. None of that makes any AI companion app appropriate for a 13-year-old: the honest answer for that age group is “none of them,” not “this one instead.”
A Safer Alternative: Built for Adults, From the Start
Parents looking for creative AI roleplay, storytelling, and character-driven games for themselves don't have to choose between an unregulated free-for-all and nothing. chatbrat.ai takes a different approach, and it's not a marketing claim. It's our written Child Safety Policy:
- Strictly 18+, by policy. chatbrat.ai is an adults-only platform. Users must be 18 or older; we use age-attestation at signup and account controls that block under-18 access. Any account we determine belongs to someone under 18 is terminated immediately, with data deleted within 48 hours per our Child Safety Policy. See also our Community Guidelines.
- Zero-tolerance CSAM policy, covering AI-generated and written content, not just photographs. We commit to NCMEC CyberTipline reporting under 18 U.S.C. § 2258A as described in our Child Safety Policy.
- Real persistent memory, so the product doesn't need to manufacture urgency or dependency to feel worth returning to.
- Creative range for adults: social-deduction and murder-mystery games, fantasy and romance characters, custom scenarios and story arcs, without an engagement-at-all-costs design aimed at teenagers.
An 18+ platform, on purpose
Read the actual policy, not just the pitch. Our Child Safety Policy is public.
Explore chatbrat.ai →Frequently Asked Questions
Is Character.AI safe for a 13-year-old?
Has Character.AI been sued over a teen's death?
Did Character.AI fix its safety problems?
Is chatbrat.ai safe for teenagers?
The Bottom Line
Character.AI was built to maximize engagement, and for a long stretch that came before minor safety, not alongside it. The guardrails that exist now (the under-18 chat ban, face verification, the settlements) are real, but they were forced by lawsuits and regulators, not offered proactively. Parents should treat that history as the baseline, not the current marketing.
Compare the full landscape of Character.AI alternatives in Best Character.AI Alternatives 2026, and read chatbrat.ai's own Child Safety Policy before deciding what's appropriate for your household.
Sources
- CNN: Character.AI and Google agree to settle lawsuits over teen mental health harms and suicides
- JURIST: Google and Character.AI agree to settle lawsuit linked to teen suicide
- NPR: Lawsuit on a chatbot that hinted a kid should kill his parents over screen time limits
- Common Sense Media: AI Companions Decoded, safety standards and risk assessment
- Texas Attorney General: Investigation into Meta and Character.AI over deceptive AI-generated mental health services
- Character.AI: Taking Bold Steps to Keep Teen Users Safe (official announcement)
- CNBC: Character.AI to ban teens from open-ended chats
Garret Williams is the founder and CEO of chatbrat.ai, building at the frontier of AI companions and roleplay chatbots. Before pivoting to tech, the Michigan native studied marketing at UCLA and directed acclaimed film projects—including his TV pilot Self-Care (nominated for Best TV Episodic at the 2023 Mammoth Film Festival) and Eco-Riot (featured on MasterClass). He writes The Bratlog to document the uncharted territory of AI relationships, sharing real-time lessons and tackling the open questions nobody has answered yet.