From the OpenAI edition of October 30, 2026
Common Sense Media rates ChatGPT for Teens "Unacceptable Risk": no parent alerts on new accounts, missed crisis referrals, bypassable study mode
Common Sense Media Youth AI Safety Institute · K-12 · United States
- Publisher
- Common Sense Media
- Original publication
- October 7, 2026
- Source retrieved
- 2026-10-31
What happened
Testing more than 4,000 prompts on teen-registered accounts, with clinician review, the Institute found zero parent alerts on newly linked accounts after up to an hour of self-harm or disordered-eating conversation, more than one in four warranted crisis referrals missed, a study mode that still completed homework, and age estimation that never moved adult-registered accounts to the teen experience.
Why it matters
Newly relevant: OpenAI's Oct. 30 safeguards and the state AG suits respond to the same gaps, so this is the main independent baseline districts can use to judge whether the new protections close them.
Evidence and measured results
Before-and-after testing around the Aug. 18 ChatGPT for Teens launch; expert review by child psychiatrists and a pediatrician; summarized in the release.
Limitations and uncertainty
Published Oct. 7, before OpenAI's Oct. 30 changes. The Institute is funded partly by the OpenAI Foundation and industry; full risk assessment not inspected. OpenAI disputed parts of the findings (the Institute's rebuttal is already in the archive).
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-10-31; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Districts cannot tell whether ChatGPT teen safeguards actually protect students.
- Stakeholders
- Student services and counselors
- Technology director
- School board
- Parents
- Discovery questions
- Does district policy rely on vendor parent alerts?
- Do students use consumer or school-managed accounts?
- Who would run an independent re-test?
- Value hypothesis
- Independent testing gives boards defensible grounds for student-access decisions.
- Potential engagement
- A bounded re-test of teen safeguards using a published-method protocol after the Oct. 30 rollout.
- Unsupported claims
- That OpenAI's new safeguards close these gaps
- That school-managed workspaces behave the same
Pre-sales engineering
Role takeaway
- Fit
- District evaluation of student-facing AI tools.
- Architecture and integration
- None; vendor-side behavior.
- Prerequisites
- Test accounts
- Clinical reviewer access
- Test prompt set
- Constraints
- Consumer accounts tested
- Pre-rollout results
- Security
- Test data involves sensitive themes; handle under research protocols.
- Proof of value
- Parent-alert test on a newly linked account
- Crisis-referral rate on a prompt sample
Delivery
Role takeaway
- Work
- Document compensating controls and schedule a post-rollout re-test.
- Dependencies
- OpenAI rollout
- Counselor capacity
- Ownership
- District student services with IT.
- Skills and adoption
- Counselor and parent guidance on not relying on built-in alerts.
- Governance checkpoints
- Board review of re-test results
- Acceptance criteria
- Re-test completed within 60 days of rollout (proposed)
- Crisis-referral miss rate reported (proposed)
- Risks
- False confidence in alerts
- Liability if failures recur
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Applies to consumer teen accounts; school-managed Edu or Teachers workspaces were not the test subject.
Governance
Who approves, reviews and stays accountable for outcomes?
Supports requiring independent testing before districts permit or recommend student ChatGPT use.
Security and privacy
What data, permissions and controls need testing?
Parent alerts and age estimation both depend on processing minors' conversation data.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Counselors may need to cover gaps left by unreliable automated alerts.
Procurement
What should contracts, pricing and exit terms secure?
Supports independent-testing clauses before student-facing AI adoption.
Operating model
Which teams own the service once it runs?
Shifts safety assurance from vendor features to district verification.
Publication history
- 2026-10-30OpenAI · Issue 027 resources
Stable resource ID: common-sense-chatgpt-for-teens-unacceptable-risk-20261007