ChatGPT for Teens: Parent Alerts That Never Arrive
A Common Sense Media evaluation found that the teen mode does not notify parents in crisis situations.
Common Sense Media rated ChatGPT for Teens as an "unacceptable risk." The organization reviewed the parent alerts in the teen mode and concluded that they do not arrive when they should.
That experience, presented in August, includes limits designed for students. The evaluation maintains that it does not send notices to parents when appropriate, does not offer the right help in a crisis, and still does kids' homework.
OpenAI rejects that diagnosis. Eric Porterfield, a company spokesperson, stated that much of the testing may have started and ended before the parental controls were complete, so its results would not establish whether the notifications work as designed.
«No app replaces human care»
The company added that it remains committed to teen safety and to giving parents useful tools to support them.
An alert that never arrives
Tom Siegel, executive director of Common Sense Media's Youth AI Safety Institute, said it bluntly: "A teenager can spend an hour talking about self-harm without their parent receiving a single alert" (Sígueme translation). Hence his conclusion: until OpenAI fixes it and proves it with independent testing, ChatGPT should be for adults only.
Siegel warned that this mode can give parents a false confidence in controls and alerts that often do not work. The underlying problem is trust: someone may believe a safety net exists when the notice never arrives.
Two readings of the same test
Siegel responded that before starting, they confirmed with OpenAI that several features of the teen mode were already active, including notifications about eating disorders and study mode.
Regarding parent alerts, the company revealed after the tests that its systems can take several hours to activate when an account is first linked. Some test accounts were linked within that window; others had been linked much longer and still did not receive notifications.
That is why the organization maintains its conclusion: parent alerts are not reliable in a crisis. The discussion is, at bottom, verifiable: when it was tested, on which accounts, and with what configuration.
What you can check at home
No app replaces human care. You can turn on teen mode and, at the same time, ask your younger sibling how their day went. The tool organizes; conversation accompanies.
Go with your parents into the family plan and review three things together: that the account is linked, that notifications are turned on, and that both of you know where they appear. According to what OpenAI responded, a newly linked account can take hours to activate its systems, so the review is not a one-time procedure.
If someone in your home is going through something serious, do not wait for the notification. Talk to a trusted adult, a health professional, or your pastor; in an emergency, call your country's emergency services.
"Train up a child in the way he should go; even when he is old he will not depart from it" (Proverbs 22:6, ESV). In a home, training means being present: asking, listening, and asking again. That is not programmed in a control panel.
This week, sit down with your parents and review the family plan alerts as a conversation about how you care for each other, not as a procedure. Technology changes every month; a family's care is built by hand. Who checks your alerts and who checks your younger sibling's?
