Quick safety summary
Character.AI at a glance
18/100
not recommended
Ages 18+
Not recommended
Low
Score 22/100
15/100
Inconsistent
Best use cases
- AI Companion / Roleplay Chatbot workflows
- Personal exploration
Main risks
- Privacy & Data
- Content Exposure
- Generative Safety
Not ideal for
Product overview
How Character.AI works
Character.AI is an open-ended AI persona chatbot platform that allowed users of any reported age to form parasocial, often romantic or sexual relationships with AI characters. Between 2024 and 2026 it became the center of the most serious documented child AI safety crisis in history, including the suicide of 14-year-old Sewell Setzer III, family lawsuits, an FTC inquiry, and a January 2026 Kentucky Attorney General lawsuit. Under legal pressure, Character.AI banned open-ended chatting for users under 18 in November 2025, but this reactive measure does not redeem the platform's record of documented harm.
Student usage
Effectively prohibited — under-18 users banned from open-ended chat as of November 25, 2025
Modality
Text chat; voice calls (c.ai+ subscribers)
Underlying model
Proprietary in-house LLM (separate under-18 variant prior to Nov 2025 ban)
Common use cases
- Roleplay and creative storytelling
- Emotional/companionship chatting
- Entertainment with fictional/celebrity bots
- Casual language practice
Free full report — 0 of 3 used
You can open 3 complete evaluations on the free plan. Already-opened reports stay unlocked. Upgrade to School Pro for unlimited full reports and the K-12 governance workspace.
Risk intelligence
Risk level breakdown
Risk dimensions derived directly from the SafeGradeAI rubric scores (higher score = higher risk). Additional dimensions appear when entered by an evaluator.
Scores derived from the SafeGradeAI rubric. Methodology: read more.
Safety framework
Framework analysis
Five evaluation categories from the SafeGradeAI rubric. Each shows the rubric score, evaluator confidence, and (where documented) analysis, strengths, and concerns.
Strengths & concerns
Detailed pros & concerns
Strengths
- Eventually enacted full open-ended chat ban for under-18 (November 2025)
- Introduced third-party age assurance (Persona) in late 2025
- Launched Parental Insights dashboard in 2024
- Added crisis intervention redirects to mental health hotlines
- Two-hour daily usage limits introduced before the full ban
Concerns
- 14-year-old Sewell Setzer III died by suicide after months of AI chatbot relationships
- Multiple families sued in 2025 alleging teen suicide, attempts, and psychological harm
- Kentucky AG filed lawsuit (January 2026) alleging predatory design
- FTC issued Section 6(b) orders in September 2025
- HEAT Report documented sexual exploitation appearing in child-age accounts
- Age verification was self-reported birthdate only until late 2025
- Platform design optimized for emotional dependency and parasocial attachment
- Common Sense Media rated 'Unacceptable' risk in April 2025
- November 2025 safety pivot was reactive — driven by lawsuits not safety design
Guidance at a glance
Recommendations for parents, teachers, and administrators
Parent guidance
Character.AI should not be used by anyone under 18, full stop. Even with the November 2025 'ban' on open-ended chat for minors, age verification can be circumvented by falsifying a birthdate. Parents should actively block the app and website on all family devices. If a child has previously used Character.AI, intimate conversation logs were collected. The platform's design — personas that mimic romantic partners combined with AI that never tires of conversation — is engineered for emotional dependency.
Classroom-use recommendation
Not recommended for any classroom use. Persona chats, weak age gating, and documented safety incidents make it inappropriate on school networks or devices; block at the network layer and address it only inside a media-literacy lesson about parasocial AI risks.
School policy recommendation
Schools and districts should place Character.AI on content filter blocklists immediately. The platform has no educational mission, no curriculum alignment, and no meaningful safety tier for students. Educators aware of students discussing Character.AI characters or relationships should treat this as a student wellbeing flag requiring counselor follow-up. Staff should not use Character.AI as an instructional demonstration tool.
Trust & credibility
Evaluation integrity
Last reviewed
June 29, 2026
Report version
v1
Evaluator confidence
19/100
Monitoring
Active continuous monitoring
Methodology
Scored against the SafeGradeAI rubric covering privacy, moderation, generative safeguards, age-appropriate design, and transparency.
Read full methodology →Change log
- June 29, 2026Evaluation published.
Re-tested at least every 6 months and after any major release.
What changed since the previous evaluation
Reviewed June 29, 2026 · Confidence: high
- 2021-11: Character.AI founded by former Google LaMDA researchers Noam Shazeer and Daniel De Freitas
- 2022-09: Public beta launched
- 2023-03: $150M Series A led by a16z at a $1B valuation
- 2024-08: Google licenses Character.AI's technology; Shazeer and De Freitas return to Google
- 2024-10-23: Megan Garcia files wrongful-death lawsuit alleging the product contributed to her teenage son's suicide
- 2025-04-10: Common Sense Media rates Character.AI 'Unacceptable' risk for teens
- 2025-09-11: FTC issues Section 6(b) orders to Character.AI and other AI companion companies
- 2025-09-16: Heat Initiative HEAT Report and multiple new family lawsuits
- 2025-10-29: Open-ended chat for users under 18 banned
- 2025-11-25: Persona-based age-assurance system rolled out
- 2026-01-07: Settlements announced with Google in teen-harm lawsuits
- 2026-01-08: Kentucky Attorney General files consumer-protection lawsuit
Sources & citations
Every major claim in this evaluation is backed by at least one credible public source. Last reviewed June 29, 2026.
- 1Garcia v. Character Technologies — ComplaintMiddle District of Florida · 2024-10-23 · cdn.arstechnica.net
- 2Character.AI and Google sued after teen's deathThe Verge · 2024-10-23 · theverge.com
- 3AI Risk Assessment: Character.AI — UnacceptableCommon Sense Media · 2025-04-10 · commonsensemedia.org
- 4FTC Launches Inquiry into AI Chatbots Acting as CompanionsFederal Trade Commission · 2025-09-11 · ftc.gov
- 5HEAT Report: Sexual Exploitation on Character.AI Kids' AccountsParentsTogether Action · 2025-09-28 · parentstogetheraction.org
- 6More families sue Character.AI over teen suicideCNN Business · 2025-09-16 · cnn.com
- 7Character.AI banning minors from AI character chatsThe Verge · 2025-10-29 · theverge.com
- 8AG Coleman Sues AI Chatbot Company for Preying on ChildrenKentucky AG · 2026-01-08 · kentucky.gov
Methodology
How this score was produced
Scored against the SafeGradeAI framework v2.5 across five dimensions — privacy, moderation, GenAI safeguards, age-appropriate design, and transparency — with COPPA, FERPA, and GDPR-K lenses. Independent. No vendor pay-for-rating.
Independent evaluation. SafeGradeAI does not accept payment from vendors for ratings.
Official website
Visit Character.AI directly
For the most current product capabilities, pricing, and policies, visit the official Character Technologies, Inc. website.
character.ai