ChatGPT for Teens is facing renewed scrutiny after new research questioned whether some of OpenAI’s safeguards for younger users work as intended.
OpenAI has introduced several protections aimed at users between 13 and 17. These include parental controls, restrictions on sexual content, limits on emotionally dependent conversations and tools designed to encourage healthier use.
However, research from Common Sense Media has raised concerns about whether some of those protections are reliable enough, particularly when conversations involve self-harm or other sensitive issues.
What Is ChatGPT for Teens?
OpenAI introduced ChatGPT for Teens as a safer version of the chatbot experience for younger users.
Teen accounts are designed to operate with additional restrictions compared with standard adult accounts.
Among the safeguards are limits on romantic or emotionally dependent interactions. The system is also designed to avoid presenting itself as conscious or encouraging teenagers to develop unhealthy emotional attachments to the chatbot.
Other protections include restrictions on sexualised images and warnings when younger users attempt to share potentially sensitive pictures.
Parents who connect their account with a teenager’s account can also access additional controls.
OpenAI Says Teen ChatGPT Use Is Mostly Educational
OpenAI says its own data suggests teenagers generally use ChatGPT for relatively short periods.
According to the company, the average teenage user spends less than 15 minutes per day using the service.
A large amount of teen activity also appears to involve learning. This can include help understanding school subjects, studying and receiving guidance on assignments.
OpenAI also uses automatic break reminders when teenagers spend extended periods using the chatbot.
The company says nearly half of younger users stop shortly after receiving such a reminder.
For teenagers who spend three consecutive hours or more on ChatGPT, OpenAI said more than 80% of their prompts during those sessions contained some connection to learning.
Study Raises Concerns About ChatGPT for Teens Safeguards
Despite those findings, Common Sense Media reached a more critical conclusion after testing the platform.
The organisation described the current system as presenting an unacceptable level of risk for children and teenagers.
Its researchers tested accounts registered to teenage users both before and after OpenAI introduced additional protections.
The study found that ChatGPT was effective at refusing certain inappropriate sexual roleplay requests.
However, researchers reported weaker results in several other areas.
One major concern involved parental notifications.
Researchers Question Parental Safety Alerts
Parental alerts are intended to provide adults with information when a linked teenager appears to be discussing subjects such as self-harm or eating disorders.
According to the research, newly created parent-linked accounts did not consistently trigger alerts during conversations involving serious mental health topics.
The researchers said alerts were more likely to appear on older accounts that had accumulated a longer history of sensitive conversations.
This raised concerns that parents could believe they will always receive immediate warnings when serious issues emerge.
OpenAI has challenged that interpretation.
The company said parts of the testing may have taken place before parental controls were completely activated. It also disputed whether the research accurately represented how the safeguards operate in everyday use.
Crisis Responses Also Come Under Scrutiny
Another concern raised by the study involved the way ChatGPT responds when teenagers discuss a crisis.
Researchers said the chatbot did not consistently direct young users toward professional support or crisis services.
That finding is particularly important because conversations involving self-harm require a much stronger level of protection than ordinary chatbot interactions.
AI companies have increasingly been expected to develop systems that recognise high-risk conversations and encourage users to seek appropriate human assistance.
For younger users, those protections carry even greater importance.
Researchers Say ChatGPT Can Still Act Like a Friend
OpenAI has attempted to reduce conversations that could encourage teenagers to think of ChatGPT as a human companion.
However, the study found that the chatbot could still sometimes use language that appeared highly personal or friend-like.
Researchers also found that the system could continue assisting with schoolwork in ways that may go beyond simple tutoring.
This creates another challenge for AI companies.
Educational assistance can help students understand difficult subjects, but there is a fine line between supporting learning and completing work on behalf of the student.
OpenAI Pushes Back Against Research Findings
OpenAI has disagreed with parts of Common Sense Media’s assessment.
The company said independent evaluation is valuable but argued that the study did not fully reflect how its teen protections work in real-world situations.
This disagreement highlights a wider debate surrounding artificial intelligence.
AI companies are rapidly adding parental controls and safety tools, while independent researchers are trying to determine whether those measures remain effective during long and complicated conversations.
As AI becomes more common in homes and schools, those tests are likely to become increasingly important.
Teen AI Safety Becomes a Bigger Technology Issue
Concerns about young people’s relationship with digital platforms have grown significantly in recent years.
Social media companies have faced increasing pressure over claims that certain online experiences can negatively affect younger users.
Artificial intelligence services are now entering the same debate.
Unlike traditional social media platforms, AI chatbots can hold lengthy and highly personalised conversations with individual users.
That creates new opportunities for education and creativity. However, it also creates new safety challenges when children use AI to discuss personal or emotional problems.
Why Reliable Parental Controls Matter
Parents may increasingly depend on built-in safety features when allowing teenagers to use AI platforms.
For that reason, alerts and restrictions need to behave consistently.
A safeguard that works only under certain circumstances could create a false sense of security.
At the same time, parental controls are only one part of the solution.
Young users also need clear guidance about what AI can and cannot do, while parents and educators need to understand that chatbots should not replace qualified professionals in serious situations.
For more developments in artificial intelligence and digital technology, readers can follow the latest Gen Z Gram technology coverage.
Conclusion
ChatGPT for Teens shows how difficult it can be to balance the educational benefits of artificial intelligence with the safety needs of younger users.
OpenAI says teenagers are mostly using ChatGPT for learning and relatively short sessions. However, independent research has raised questions about parental alerts, crisis responses and emotionally personal conversations.
The disagreement is likely to fuel further scrutiny of how AI companies protect minors.
As more teenagers begin using AI for school, advice and everyday questions, reliable safeguards will become increasingly important for parents, educators and technology companies.
