OpenAI has rolled out a dedicated teen experience for ChatGPT, automatically applying a set of default safety protections and study tools to accounts belonging to 13-to-17-year-olds. The launch, announced August 18, 2026, consolidates more than a year of piecemeal safety work — parental controls, age-prediction technology, and updated content rules — into a single experience that activates without any action required from the teen or their parents.
The timing matters. OpenAI is rolling this out while facing multiple wrongful-death lawsuits alleging that ChatGPT contributed to harm involving minors, and while an ongoing Federal Trade Commission inquiry examines how AI chatbot companies protect children and teens. Neither the lawsuits nor the FTC inquiry has been resolved, and OpenAI has not linked the launch directly to either — but the backdrop is impossible to ignore in a product aimed squarely at making the chatbot safer for young users.
Table of Contents
What Is ChatGPT for Teens?
ChatGPT for Teens is not a separate app or a paywalled tier. It’s a mode within the standard ChatGPT product that switches on automatically when OpenAI’s systems determine a user is between 13 and 17 — either because the account holder states that age at sign-up, or because the company’s age-prediction system estimates the user is under 18.
The rollout began globally on August 18, 2026, and is available on both Free and paid personal ChatGPT plans, according to OpenAI’s help documentation. Full availability in Australia is expected by September 8, 2026, and OpenAI says it will continue expanding the feature “over the coming weeks,” suggesting some regions may see a staggered rollout beyond what the company has specified.
The launch responds to a scale of teen usage that has grown quickly. Pew Research has found that the share of U.S. teens using ChatGPT for schoolwork roughly doubled year over year, and a 2025 Common Sense Media study found that more than 70% of U.S. teens have used AI chatbots for companionship, with about half doing so regularly.
How the Safety Protections Work
At the core of the teen experience are what OpenAI calls “Under-18 Principles,” built into its Model Spec — the internal document that governs how its models are supposed to behave. For teen accounts, these principles translate into tighter default restrictions around self-harm, violence, eating disorders, dangerous activities, and sexually explicit or graphic content.
OpenAI has also added rules aimed at limiting how human the chatbot can seem to a teenager. Under the updated Model Spec, ChatGPT is instructed not to use romantic language, encourage emotional dependence, or imply that it has feelings or consciousness when talking with teen users. The company frames this as an effort to reduce anthropomorphism and keep teens oriented toward real-world relationships rather than a chatbot.
Teens using the mode also see more frequent break reminders during long sessions, along with prompts that identify ChatGPT as an AI tool. Before certain image uploads, a reminder encourages teens to check for sensitive information before sharing.
One feature — “responsible homework reminders” — is designed to recognize when a teen appears to be trying to shortcut an assignment rather than understand it, and redirect them toward guided study tools instead. OpenAI has not disclosed the technical method behind this detection, so it’s not possible to say how the system identifies attempted shortcuts, or how reliable that detection is in practice.
Study Mode and Learning Tools
The education side of the launch centers on Study Mode, which OpenAI describes as using guiding questions and step-by-step support to help teens work through material rather than simply handing them answers. The company says this approach draws on established learning-science research showing that active engagement and self-explanation tend to support better retention than passive review.
Around Study Mode, OpenAI has built several supporting features: quizzes, “Learning Visualizations” for difficult concepts, and Study Hours, which let a teen or a linked parent choose windows of time when new chats default into study mode automatically. The company has also announced a partnership with CodeAI, an education-focused organization, aimed at helping teens understand how AI systems work rather than just how to use them.
OpenAI cites its own early evaluations showing what it calls “promising gains in student performance” from tools like Study Mode. That claim comes from the company itself — no independent, third-party study of Study Mode’s educational impact was available at the time of this launch, so it should be read as a company claim rather than a verified outcome.
Parental Controls — What Parents Can and Can’t See
For families, the most consequential detail may be the one that limits, rather than expands, parental access. Parents and guardians can link accounts with a teen — provided the teen accepts the invitation — and from there manage certain settings, set Quiet Hours that block access during scheduled times, and receive safety notifications in what OpenAI describes as “limited high-risk situations.” The company says it is now including notifications related to eating disorders alongside its existing alerts.
What parents cannot do is read their teen’s conversations. OpenAI states this explicitly in its own help documentation: parental controls do not grant access to chat content, and if the company does send a safety notification, it shares only the information needed to support the teen’s safety — not the underlying conversation. Either the teen or the linked parent can unlink the accounts at any time, which ends the arrangement. And if OpenAI’s systems later determine an account holder has turned 18, teen protections and any linked parental controls can lapse automatically, though certain protections can be reinstated.
This is a meaningful trade-off. It preserves a degree of privacy for teenagers, but it also means parents relying on ChatGPT’s built-in tools get considerably less visibility than they might expect from consumer parental-control software built for other purposes.
Why Now: Legal and Regulatory Backdrop
ChatGPT for Teens doesn’t arrive in a vacuum. OpenAI is currently facing lawsuits alleging that interactions with ChatGPT contributed to the death or harm of young users, including a widely reported case involving a 16-year-old in California whose family sued the company in August 2025. Those cases remain in litigation, and the allegations have not been proven in court.
The past year traces a clear buildup to this launch:
- September 2025: OpenAI announces age-prediction technology and launches parental controls; the FTC opens its Section 6(b) inquiry into AI chatbots and minors.
- October 2025: California signs SB 243, regulating companion chatbots for minors.
- November 2025: Character.AI bans open-ended chat for users under 18.
- August 2026: OpenAI launches ChatGPT for Teens.
Separately, the FTC opened a formal inquiry in September 2025 into how seven companies — including OpenAI, Meta, Google, Snap, and Character.AI — evaluate the safety of chatbots that can function as companions for children and teens. That inquiry is still active, and as of this launch, the FTC has not published findings or conclusions.
The launch also comes as OpenAI pursues a public listing, following a March 2026 funding round that valued the company at $852 billion. OpenAI has not linked the teen launch to its IPO plans, and the two should be understood as concurrent rather than connected by any stated company rationale.
On the legislative side, California’s SB 243, a law regulating companion chatbots for minors, took effect January 1, 2026. At the federal level, a proposal known as COPPA 2.0 would extend privacy protections currently limited to children under 13 up through age 17 — but that bill has not passed Congress, and existing federal law (COPPA) still applies only to users under 13.
OpenAI has not stated that the teen launch was designed in direct response to any specific lawsuit or the FTC inquiry. The company frames it instead as the next step in safety and education work it has been building for more than a year. Readers can draw their own conclusions about the relationship between the legal pressure and the timing, but the two are best described as concurrent rather than causally linked based on available evidence.
How It Compares to Other AI Chatbots for Teens
The landscape of AI chatbot safety for minors varies considerably by company, and no platform has been independently certified as the definitive safest option.
| Platform | Minimum Age | Parental Visibility Into Chats | Independent Risk Rating (Common Sense Media, prior assessment) |
|---|---|---|---|
| ChatGPT for Teens | 13–17, auto-applied | Alerts only, no transcript access | Not yet separately rated |
| Snapchat My AI | 13+ (Snapchat’s general minimum) | Activity visibility only, no chat content | Not separately rated in sources reviewed |
| Character.AI | 18+ (open-ended chat) | N/A — under-18 chat unavailable | Not separately rated in sources reviewed |
| Google Gemini (teen experience) | Teen-specific mode | Not detailed in sources reviewed | “Low risk” |
| Anthropic Claude | 18+, no minor pathway | N/A — not available to minors | “Minimal risk” |
Character.AI took the most restrictive approach among major platforms, banning open-ended chat for all users under 18 as of November 2025, following its own wrongful-death litigation. Snapchat’s My AI, built into a platform already popular with teens, gives parents only limited visibility — they can see whether a teen has interacted with My AI through Snapchat’s Family Center, but not the content of those conversations, and the chatbot retains conversation data by default unless a user manually deletes it. Anthropic’s Claude takes a different approach entirely: it requires users to be 18 or older, with no parental-consent pathway for younger teens at all.
Independent risk assessments have varied by platform and by date. Common Sense Media, a nonprofit that evaluates media and technology for families, has in past assessments rated Google’s teen-specific Gemini experience as “low risk” and Claude as “minimal risk,” while rating standard ChatGPT as “high risk” overall for teen use — a rating issued before this specific teen-focused launch. It’s unclear whether or when that organization will re-evaluate ChatGPT for Teens as a distinct product.
Given this fragmented picture, ChatGPT for Teens should be understood as one of several different strategies for handling underage users — more restrictive than Snapchat’s approach, less restrictive than Character.AI’s outright ban, and not yet independently benchmarked against its own predecessor.
What’s Still Unknown or Unproven
Several open questions remain about the newly launched product. No independent organization has yet tested ChatGPT for Teens specifically, as opposed to standard ChatGPT — meaning claims about its safety performance rest largely on OpenAI’s own account for now. The technical mechanism behind the homework-cheating detection feature has not been disclosed. OpenAI has not published a data-retention or model-training policy specific to teen accounts, separate from its general ChatGPT privacy policy, and its compliance approach under the EU’s child-focused privacy framework (often referred to as GDPR-K) was not documented in available sources.
There’s also a practical question about how effective any of these protections can be in the real world. TechCrunch, which broke the story, noted that it remains unclear how difficult these systems will be for motivated teens to work around — teenagers have historically been adept at circumventing parental controls, and creating a new account with a different stated age remains a straightforward workaround on most platforms, including this one, according to third-party reviews of ChatGPT’s existing controls.
Finally, OpenAI has previously acknowledged — in the context of past incidents — that its safety training can become less reliable during long, extended conversations, as the back-and-forth grows and parts of the model’s guardrails degrade. Whether the new teen protections hold up better under that kind of sustained interaction hasn’t yet been independently tested.
Conclusion
ChatGPT for Teens represents a real expansion of OpenAI’s safety infrastructure: stronger default content restrictions, education-focused tools built around active learning, and a more visible parental-controls system than the company had before. Those are concrete, verifiable changes.
At the same time, the launch leaves significant gaps unaddressed by independent verification. The core promise of parental oversight is narrower than it might first appear — parents get alerts, not access. The technology behind key safety features, including cheating detection, remains undisclosed. And the product enters a legal and regulatory environment that is still actively unresolved, with lawsuits pending and a federal inquiry ongoing.
What happens next will likely say more than the launch announcement itself: whether independent safety evaluators like Common Sense Media reassess the product, whether the FTC inquiry produces public findings, and whether the courts reach conclusions in the pending litigation. Until then, ChatGPT for Teens should be understood as a significant but unproven step — one shaped as much by the legal and public pressure surrounding AI and minors as by the technology itself.
Frequently Asked Questions
What age group does ChatGPT for Teens apply to? It applies to users aged 13 to 17, based on either self-reported age at sign-up or OpenAI’s age-prediction system estimating the user is under 18.
Do parents have to opt in, or is it automatic? The teen protections themselves are automatic — no setup is required. Parental controls, however, are optional and require both the parent and teen to agree to link their accounts.
Can parents read their teen’s ChatGPT conversations? No. OpenAI’s own documentation confirms that parental controls do not allow parents to read or monitor a teen’s conversations. In limited high-risk situations, OpenAI may send a safety notification, but it shares only the information needed to support the teen’s safety, not the conversation itself.
Is ChatGPT for Teens free, or does it require a paid plan? It’s available on both Free and paid personal ChatGPT plans — it is not restricted to a premium tier.
How is ChatGPT for Teens different from standard ChatGPT? It applies the same underlying ChatGPT model but adds default content restrictions in sensitive areas (self-harm, violence, eating disorders, sexual content), education-focused tools like Study Mode, break reminders, and optional parental controls not present in the standard adult experience.
Can a teen bypass ChatGPT for Teens’ protections? Reporting on the launch, including from TechCrunch, notes that it’s unclear how resistant the system is to workarounds, and third-party reviews of OpenAI’s existing controls have found them bypassable — for example, by creating a new account with a different stated age.
How does ChatGPT for Teens compare to Snapchat’s My AI or Character.AI? The approaches differ significantly. Character.AI banned open-ended chat for users under 18 entirely in November 2025. Snapchat’s My AI offers parents only limited visibility, without access to conversation content. ChatGPT for Teens sits between these approaches, offering more structured safety and study tools than My AI while remaining more permissive than Character.AI’s outright restriction.
Sources: OpenAI (openai.com/index/chatgpt-for-teens/ and help.openai.com/en/articles/20001421-chatgpt-for-teens), TechCrunch, Associated Press, CNN, Axios, the Federal Trade Commission, and Common Sense Media, as cited throughout.
References
- OpenAI. “Introducing ChatGPT for Teens.” OpenAI Blog, August 18, 2026. https://openai.com/index/chatgpt-for-teens/
- OpenAI Help Center. “ChatGPT for Teens.” Updated August 18, 2026. https://help.openai.com/en/articles/20001421-chatgpt-for-teens