Quick safety summary
ChatGPT at a glance
52/100
not recommended
Ages Official minimum age 13; ages 13–18 require parental consent. SafeGradeAI recommendation: supervised teen use only
Not recommended
Low
Score 50/100
46/100
Inconsistent
Best use cases
- General-Purpose AI Assistant workflows
- Personal exploration
Main risks
- Privacy & Data
- Content Exposure
Not ideal for
Product overview
How ChatGPT works
ChatGPT (consumer Free and Plus tiers) is OpenAI's flagship general-purpose AI chatbot. While OpenAI made meaningful safety improvements since mid-2025 — parental controls (Sept 2025), safety-routing, a 170-expert mental-health update (Oct 2025), and Trusted Contact (May 2026) — those changes came reactively, following wrongful-death lawsuits involving minors. The consumer product still relies on self-reported age, trains on conversations by default, and was designed for adults; it is not recommended for children or unsupervised teens.
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
- Parental controls with account linking and safety notifications (Sept 2025)
- Safety routing system routes sensitive conversations to stronger models
- October 2025 mental-health update developed with 170+ experts
- Trusted Contact feature (May 2026) enables real-world support escalation
- Training opt-out available in Data Controls settings
- Model Spec publicly published
- Free tier accessible without subscription
Concerns
- Parental controls do not give parents access to teen conversation transcripts
- The "Improve the model for everyone" setting is on by default on consumer accounts and must be turned off manually
- Consumer tiers have no district admin console or contractual data terms
- General-purpose product; safety behaviour differs between Free, teen, and Edu/Enterprise contexts
- SafeGradeAI has not published independent filter-bypass testing
Guidance at a glance
Recommendations for parents, teachers, and administrators
Parent guidance
Not for children under 13. For teens aged 13–18, parental consent is required by OpenAI's terms: link a parental-control account, reduce sensitive content, turn off "Improve the model for everyone", and set quiet hours. Expect no transcript visibility — parental controls manage settings, not access to your teen's conversations, so keep talking about what they use it for.
Classroom-use recommendation
If your district has not signed a ChatGPT Edu/Enterprise agreement, do not direct students to consumer ChatGPT. Where teen accounts are used at home, teach source-checking and disclosure of AI assistance, and never ask students to paste personal or classmate information into prompts.
School policy recommendation
The consumer tiers are not appropriate as a district deployment. ChatGPT Edu, Business or Enterprise agreements provide the admin controls, retention terms and contractual data commitments a district needs. Do not extend a consumer-tier assessment to Edu, or the reverse.
Trust & credibility
Evaluation integrity
Last reviewed
August 24, 2026
Report version
v1
Evaluator confidence
53/100
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
- August 24, 2026Evaluation published.
Re-tested at least every 6 months and after any major release.
What changed since the previous evaluation
Reviewed August 24, 2026 · Confidence: high
- 2022-11-30: ChatGPT publicly launched by OpenAI on GPT-3.5
- 2023-03: GPT-4 released; ChatGPT Plus subscription launched
- 2023-05: ChatGPT iOS app released
- 2024-05: GPT-4o (omni) released with native voice and vision
- 2024-09: o1 reasoning models released in preview
- 2025-05: Memory and improved parental controls announced
- 2025-08-26: Raine family files first major wrongful-death lawsuit against OpenAI
- 2025-09-29: Parental controls and safety-routing launched
- 2025-10-27: Mental-health update released with input from 170+ clinicians; Model Spec published
- 2025-11: OpenAI faces five wrongful-death lawsuits; files defense in Raine case
- 2026-05-07: Trusted Contact feature launched for crisis routing
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.
Other general-purpose AI assistant tools we've reviewed
Independent evaluation. SafeGradeAI does not accept payment from vendors for ratings.
