Common Sense Media Says ChatGPT for Teens Falls Short on Safety

Common Sense Media says OpenAI’s ChatGPT for Teens is an unacceptable risk for young users, arguing that the product’s parental alerts and crisis safeguards do not reliably work. The nonprofit’s assessment, released Wednesday, matters because it directly challenges OpenAI’s pitch that the teen-focused version of ChatGPT is safe enough for students and families.

The criticism lands just months after OpenAI introduced the teen experience in August, positioning it as a way for minors to use ChatGPT with stronger protections and more family oversight. Common Sense Media says the safeguards still fall short in practice, while OpenAI insists the group’s testing did not accurately reflect how the system works.

What Common Sense Media found

Common Sense Media, which reviews technology and entertainment through a child-safety lens, said its evaluation found serious gaps in ChatGPT for Teens. According to the organization, the product failed to notify parents when it should have, did not provide the right support during crisis-related conversations, and still helped students complete homework in ways the nonprofit views as inappropriate for a teen safety product.

The group’s conclusion was blunt: the experience gives parents a false sense of security if the protections they are told to trust do not consistently activate. That is especially concerning, Common Sense Media argued, in situations involving self-harm, eating disorders, or other sensitive topics where delayed or missing alerts could leave caregivers unaware of a teen’s risk.

Common Sense Media’s youth AI safety team said a teenager could spend an hour discussing self-harm without a parent receiving a single alert, and it called for OpenAI to prove the reliability of the safeguards through independent testing before the product is considered appropriate for teens.

Why the report matters

This dispute goes beyond one product review. ChatGPT is one of the most visible consumer AI tools in the world, and the company has spent the past two years under intense pressure from parents, educators, lawmakers, and safety advocates to show that generative AI can be used responsibly by minors.

Teen-focused AI products are now being judged not only on usefulness, but on whether they can reliably route the hardest cases to adults. That standard is far higher than a normal consumer app review. A missed notification in a homework app is one thing; a missed notification in a crisis context is far more serious.

Common Sense Media’s findings add to the broader debate about whether AI chatbots can be made safe for children through prompts, labels, and dashboards alone. Critics argue that these systems are still too easy to misunderstand, too inconsistent in edge cases, and too likely to be treated like a substitute for human guidance.

How ChatGPT for Teens is supposed to work

ChatGPT for Teens was introduced as a version of the product designed with additional guardrails for younger users. OpenAI has said the goal is to help students learn while giving parents more meaningful controls over their children’s use of the chatbot.

In principle, the teen experience includes protections intended to respond to sensitive topics, reduce risky interactions, and support family oversight. OpenAI has also framed the tool as a way for teenagers to benefit from AI without exposing them to the full, less restricted version of the product.

The problem raised by Common Sense Media is not the concept of safeguards, but their reliability. A safety system that only works sometimes can be worse than one that is clearly absent, because families may believe they are protected when they are not.

Safety features at issue

Common Sense Media highlighted three broad areas of concern:

• parent notifications that did not consistently arrive when expected;
• crisis-response behavior that the organization said was inadequate; and
• academic assistance that still amounted to homework completion.

Those concerns go to the heart of how teen AI products are judged. Parents want visibility, schools want academic integrity, and safety groups want a clear response path when a young user appears distressed. The report suggests OpenAI has not yet met that bar, at least in Common Sense Media’s testing.

What OpenAI says in response

OpenAI strongly disputes the assessment. Company spokesperson Eric Porterfield said the company remains committed to teen safety and to building tools that give parents real control over how their teens use AI.

OpenAI said it welcomes outside review, but argued that Common Sense Media’s tests did not accurately capture how the teen safeguards operate in practice. The company’s main objection is procedural: it says much of the testing may have begun and ended before parental controls were fully activated on newly linked accounts.

If that timeline is correct, OpenAI argues, the results would not prove whether safety notifications failed as designed. The company said it wants to keep working with outside experts so teens can use the product for learning and creativity while parents retain confidence in the system.

OpenAI said it believes the nonprofit’s methodology may have captured accounts before all parental controls had fully taken effect, and argued that this could have affected the findings about whether alerts were working as intended.

How did Common Sense Media defend its testing?

Common Sense Media stood by both its methods and its conclusions. The nonprofit said it checked with OpenAI before testing to confirm that Teen mode features, including eating-disorder notifications and study mode, were fully launched.

On parental notifications specifically, the organization said OpenAI disclosed only after the tests that its systems may take several hours to activate on newly linked accounts. Common Sense Media said some of its test accounts fell inside that window, but others remained linked for much longer without receiving alerts, which it says shows the notifications are still unreliable in crisis scenarios.

The group also said it shared its findings with OpenAI before making them public, in an effort to give the company a chance to improve the product. It urged OpenAI to address the underlying safety issues rather than question the legitimacy of the evaluation.

What this fight says about teen AI safety

The disagreement highlights one of the central tensions in consumer AI: companies want to market powerful tools for broad audiences, including children and teenagers, while safety advocates want systems that behave predictably under stress. Those goals are not always compatible.

For now, the question is not whether OpenAI can describe safety features on paper. It is whether those features work consistently enough to justify use by minors, especially in situations involving self-harm, eating disorders, emotional distress, or academic misuse.

The clash also shows the importance of independent testing. AI companies often publish safety claims based on internal evaluations, but child-safety advocates argue those claims need outside scrutiny. Without it, parents may have little way to judge whether a teen-focused mode is genuinely protective or only cosmetically safer.

Why parents are watching closely

Parents are likely to care less about the branding of a teen mode and more about three practical questions:

1. Can the system reliably warn adults when something dangerous comes up?
2. Does it avoid helping teens bypass schoolwork in ways that undermine learning?
3. Does it handle sensitive conversations with the caution families expect?

Common Sense Media’s answer is no, at least not reliably enough. OpenAI’s answer is that the safeguards are working in practice and that the tests did not fairly measure them. Until those claims are reconciled, parents are left with conflicting narratives from two organizations with very different perspectives but a shared stake in child safety.

Timeline of the dispute

The debate unfolded quickly after the teen-focused feature’s launch. The key steps are below.

What happens next?

The immediate next step is pressure, not resolution. OpenAI will likely face continued scrutiny over how quickly its parental tools activate, how reliably alerts are sent, and whether the teen experience truly behaves differently from the standard chatbot when it matters most.

Common Sense Media, meanwhile, has signaled that it sees independent evaluation as essential before families can trust a teen-oriented AI product. That means the company’s response may not be enough on its own unless future testing confirms the safeguards work under realistic conditions.

For OpenAI, the stakes are broader than one nonprofit review. Teen safety has become a benchmark issue for the entire AI industry. If a leading chatbot cannot convince child-safety experts that its protections are dependable, the company may find that regulators, schools, and parents become even more skeptical of AI products aimed at young people.

For now, the message from the latest assessment is clear: a teen mode is only as good as its weakest safety link, and in Common Sense Media’s view, ChatGPT for Teens still has too many of them.

Key facts at a glance