[Audio version coming soon: this article will have a listen option at the top.]

By Janice Robinson-Celeste

A teen can have the most honest conversation of their week with ChatGPT at 11 p.m., and your phone may never buzz. That is the key gap to understand about ChatGPT parental controls: linking accounts does not mean you’ll know what your teen says or receive an alert every time a conversation worries you.

ChatGPT parental controls: a phone face down on a deep green duvet beside school notebooks, wired earbuds and a glass of water in warm nighttime light

Quick Review

  • Oregon’s November 17 lawsuit alleges that ChatGPT for Teens safeguards can be bypassed and parents were not alerted to certain risky conversations.
  • OpenAI parental notifications are limited and are not real-time monitoring.
  • A parent cannot see a teen’s chats or conversation history through linked accounts.
  • Teens can unlink accounts; notifications then stop. Teen protections and controls also end when an account leaves the teen experience at 18.
  • Use controls as one layer of support, alongside regular, non-punitive conversations and trusted human help.

What Oregon Alleges

Oregon Attorney General Dan Rayfield filed a lawsuit against OpenAI on November 17 in Multnomah County Circuit Court, seeking at least $2.035 billion and court-ordered safeguards. The state alleges violations of Oregon’s Unlawful Trade Practices Act, negligence, unjust enrichment, and public nuisance. Those are allegations, not proven facts. No court has ruled on the case.

The complaint specifically alleges that ChatGPT for Teens, which can link a parent’s account to a teen’s, has safeguards that are easily bypassed. It also alleges parents were not alerted when young users asked about suicide, self-harm, eating disorders, or mental-health issues. Oregon further alleges that age prediction is imperfect and that ChatGPT’s human-like design can encourage emotional dependence.

Rayfield said, “OpenAI knows that ChatGPT is harming children, and has failed to meaningfully fix their dangerous platform.” He also said the lawsuit did not stem from one specific Oregon incident. Read the Oregon Department of Justice announcement and OPB’s report on the allegations and company response.

Oregon’s case follows the coordinated lawsuits that eight states filed against OpenAI in October, which SCN covered in States Sue OpenAI Over ChatGPT Harms to Kids.

This lawsuit is part of Oregon’s “Keeping Kids Safe” initiative, which also includes a $125 million settlement with Meta, an investigation into Discord, and expansion of the state’s Internet Crimes Against Children task force.

What OpenAI Says

OpenAI spokesperson Drew Pusateri said the company is committed to AI policies “that apply to the entire AI industry and not just one company.” OpenAI has pointed to parental controls, break reminders, and age-appropriate model protections that exist today, and has said keeping communities safe is a top priority. The company disputes the allegations; the court has not decided whether they are true.

OpenAI’s parental-controls documentation says linked accounts let a parent manage selected settings and receive safety notifications in limited situations. It also states plainly that the system is not real-time monitoring, may miss concerns, and does not replace professional care or emergency services. Earlier SCN coverage of ChatGPT for Teens explains more about the teen experience.

What ChatGPT Parental Controls Will Alert You About

OpenAI says alerts may be sent in specific situations, such as when specially trained reviewers identify a serious self-harm concern or behavior that may be associated with disordered eating. Violence-related notices may apply when a conversation is ended or an account is temporarily suspended or permanently closed for serious concerns. Some of these notifications are gradually rolling out and may not be available on every account.

An alert does not include a transcript, chat history, or details of what your teen said. And a linked account does not give you access to real-time activity. OpenAI’s October 30 safety update describes expanded violence-related protections and says the company is working on additional notifications.

What The Settings Actually Control

It helps to separate two very different things inside ChatGPT parental controls. One is access, meaning what your teen can use. The other is information, meaning what you are told.

On access, a linked parent can turn Voice Mode, image generation, saved memories, the Health feature and the cloud browser on or off, and can decide whether your teen’s conversations may be used to improve the models. Quiet hours are the strongest tool in the set, because they can block ChatGPT entirely during scheduled blocks, including ChatGPT Work. Study hours work differently: they start eligible new conversations in study mode, but they do not block access, so they are a routine, not a lock.

On information, the list is short. OpenAI says a notification may come after specially trained reviewers identify a serious self-harm concern, after reviewers identify behavior that may be associated with disordered eating, when an account is permanently closed for conversations about acts of violence, or when a conversation about violence or harm to others is ended or the account is temporarily suspended. OpenAI notes that some of these notices are still rolling out and may not reach every account yet.

That imbalance is the part parents miss. You can shape the settings fairly easily. You cannot count on being told what worries you most.

A Black father and teenage son sit side by side at a kitchen island in late afternoon light, talking while a phone rests face down between them

What The Alerts Never Cover

The system does not promise an alert for every troubling or sensitive conversation. A teen asking about sadness, loneliness, a breakup, bullying, school stress, body image, or what self-harm means is not automatically the same as a serious concern identified by a trained reviewer. The Oregon complaint alleges that parents were not alerted to risky queries about suicide, self-harm, eating disorders, and mental health. That claim remains an allegation, not a court finding.

There are also clear circumstances when a parent should not expect alerts: if accounts were never linked, if the teen unlinks, or after the account leaves the teen experience at 18. The invitation itself is voluntary, and a teen who never accepts one leaves a parent with no controls and no notices at all. OpenAI says either person can unlink at any time. The parent is notified if the teen unlinks, but then loses the ability to manage settings or receive safety notifications. Its age-prediction system uses behavioral and account signals and can be wrong.

The alert you are hoping for exists only in a narrow set of situations, and a teen can unlink the account.
Know a parent who needs this? Send it to them.

Five Things To Set Or Say Tonight

  1. Link accounts together. In ChatGPT settings, review parental controls and confirm the teen accepts the invitation. OpenAI says linking does not provide access to chats.
  2. Schedule quiet hours. Set a block around sleep and school if that fits your family’s needs. Quiet hours block ChatGPT during scheduled times; study hours do not block access.
  3. Talk about the unlink button before it happens. Explain that the goal is safety and trust, not secret surveillance, and discuss what happens to alerts if the connection ends.
  4. Decide together what a serious concern looks like. Make a family plan for when a teen should come to a trusted adult rather than rely on a chatbot.
  5. Keep a human on call. Choose trusted adults and professional resources your teen can reach without fear of punishment. Oregon DOJ also advises parents to ask what kids discuss with AI and not to assume controls or age checks catch everything. See SCN’s guide to parental controls that can fail and Parents’ AI Emergency Guide.

Try this opener: “I’m not here to read your chats or get you in trouble. I want to understand how you use ChatGPT. What do you ask it about, and has it ever said something that felt confusing or too personal?”

If A Safety Notification Arrives

Check in calmly and without punishment. Ask whether your teen is safe right now, listen, and help them reach a trusted adult or qualified professional. A notification is a reason to connect, not a substitute for checking directly. In the United States, call or text the 988 Suicide & Crisis Lifeline; for crisis support in other countries, use Find A Helpline. If anyone may be in immediate danger, contact local emergency services.

Pinterest graphic: ChatGPT won't always tell you, what ChatGPT parent alerts miss

Frequently Asked Questions

Will ChatGPT Tell Me If My Teen Is Struggling?

Not reliably or in every situation. OpenAI says notifications apply to limited concerns and may not identify every risk. They are not live monitoring, and they do not show you the conversation.

Can My Teen Turn Off ChatGPT Parental Controls?

A teen can unlink the accounts at any time. OpenAI says the parent will be notified, but can no longer manage settings or receive safety notifications once the accounts are unlinked.

Can I Read My Teen’s ChatGPT Conversations?

No. Linking accounts does not provide access to conversations, chat history, or real-time activity. Alerts do not include transcripts.

What Happens When My Teen Turns 18?

When ChatGPT identifies a linked teen as 18 or older, the account leaves the teen experience, parental controls are removed, and teen safety notifications stop.

Have you ever asked your teen what they tell a chatbot?

If this helped, share it with another parent and follow Secure Children’s Network on Facebook, X and LinkedIn for safety alerts.

Join the Secure Children’s Network newsletter for practical resources and online-safety updates.


Discover more from

Subscribe to get the latest posts sent to your email.

Leave a comment

Discover more from

Subscribe now to keep reading and get access to the full archive.

Continue reading

Secure Children's Network logo
Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.