OpenAI has rolled out a version of ChatGPT built specifically for users under 18, adding stricter content rules and a feature that can notify a parent when the system detects signs of serious distress in a teenager’s conversation. The rollout has drawn a mixed response, with child-safety researchers welcoming the intent behind the feature while questioning whether it can reliably catch the situations it is meant to address.
How ChatGPT for Teens Detects a User’s Age
The teen-specific experience is designed to activate automatically once the system determines, through a combination of account information and behavioral signals, that a user is likely under 18. Rather than relying solely on a self-reported birthdate, which is easy to falsify, OpenAI has said it uses broader signals to make that determination, then applies the more restrictive experience by default. Age-estimation systems of this kind are an active area of research across the tech industry, and none of the current approaches are considered foolproof, since a determined teenager can still find workarounds by using a different account or misrepresenting basic details.
What Changes Once the Under-18 Rules Apply
Once a conversation is flagged as belonging to a minor, the system applies tighter content boundaries around subjects such as graphic violence, sexual content, and material related to self-harm, along with adjustments meant to reduce exposure to content the company considers inappropriate for younger users. According to reporting from The Hill, the changes are framed by OpenAI as a default state for identified teen accounts rather than an optional setting a parent has to configure separately.
How the Parental-Alert Feature Is Supposed to Work
The most notable addition is a mechanism that can flag a conversation to a linked parent or guardian account if the system detects language suggesting the teenager may be in acute distress, including indications of potential self-harm. The alert is intended to function as a safety net in situations where a young person might not otherwise reach out for help, giving a parent visibility into a crisis moment they would not normally see. OpenAI has described the feature as one part of a broader set of protections rather than a comprehensive monitoring tool that reviews every message a teenager sends.
Why Child-Safety Experts Question Its Effectiveness
Specialists who study online safety for minors have raised doubts about how consistently a language model can be expected to identify genuine crisis signals without generating false positives that erode trust or false negatives that miss real danger. Reporting from Engadget quotes experts who describe the feature as a step that looks reassuring in a press release but has not been independently tested at the scale needed to know how well it performs across the huge range of ways distress can show up in conversation. Some also point out that a teenager aware of the monitoring may simply become more guarded in what is typed, potentially undermining the very disclosures the feature is meant to catch.
The Broader Debate Over Age Verification in AI Chatbots
The skepticism sits inside a larger, ongoing argument about how AI companies should handle underage users at all, following incidents at multiple chatbot platforms that drew scrutiny over interactions with minors. Age-estimation technology, parental-consent frameworks, and content moderation for AI systems remain unsettled areas without a single accepted industry standard, and regulators in several states have begun examining whether existing child-safety laws written before generative AI existed adequately cover chatbot interactions. That regulatory uncertainty is part of why child-safety advocates tend to treat a single company’s new feature as one data point rather than a settled solution, arguing that independent audits and outside data on real-world performance will matter more than the feature’s initial description.
The parental-alert system is not automatic for every teenager on the platform; it generally depends on a parent or guardian linking their own account to a teenager’s account first, a step that requires cooperation from both sides rather than a setting a parent can enable unilaterally. That design choice mirrors similar features on other platforms, where the effectiveness of a safety tool ends up limited by how many families actually set it up, rather than by the technology itself. A teenager who does not want a parent to have that visibility can simply decline to link the accounts or use a separate account altogether, which is one of the practical gaps critics point to when questioning how much real-world protection the feature delivers.
The teen product arrives alongside a broader set of safety commitments the company has made following public scrutiny of how chatbots handle vulnerable users, including steps such as routing certain sensitive conversations to more carefully tuned model behavior and expanding crisis-resource messaging within the chat interface itself. Framing a single feature within that larger pattern helps explain why child-safety researchers are withholding judgment rather than issuing an outright dismissal, since a track record built over months of real usage data will matter more to them than the initial announcement.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
- Consumer Reports names the 2026 models it expects to break down the most
- Supplements now rank as the fifth-leading cause of death from liver disease.
- The NSA is again telling phone owners to switch off one location setting
- Four U.S. startups fired up their first small nuclear reactors, aiming to power AI data centers on-site
