What parents should know about ChatGPT for Teens

OpenAI launched ChatGPT for Teens, a version of its AI chatbot designed as a learning partner for students, with limits on emotional and intimate exchanges to address concerns about teen well-being. However, experts urge caution, saying independent testing is needed to verify its safety and effectiveness. Robbie Torney, head of AI and digital assessments at the Youth AI Safety Institute at Common Sense Media, told Mashable that while the teen experience makes important safety commitments, evidence is needed before parents put real trust in it.

What parents should know about ChatGPT for Teens Rebecca Ruiz /author/rebecca-ruiz Read Full Bio /author/rebecca-ruiz Chase DiBenedetto /author/chase-dibenedetto Read Full Bio /author/chase-dibenedetto Table of Contents This week's launch of ChatGPT for Teens https://mashable.com/tech/chatgpt-for-teens may be welcome news to parents concerned about the role of artificial intelligence https://mashable.com/category/artificial-intelligence in their child's life. OpenAI https://mashable.com/category/openai positioned its new product as a learning partner for students. The company said ChatGPT for Teens limits the emotional and intimate exchanges between young people and AI that have driven concerns about well-being and mental health. In some tragic cases, that dynamic has allegedly led teens and young adults to attempt or die by suicide after heavy engagement with ChatGPT, according to lawsuits filed against OpenAI https://mashable.com/tech/florida-first-state-to-sue-open-ai-over-links-to-violence-and-self-harm . So parents who know their teen is using ChatGPT, or who want their child to safely access AI, may see OpenAI's teen product as the answer. Yet experts who spoke to Mashable about ChatGPT for Teens urge parents to remain cautious. They say more independent testing is needed before concluding that ChatGPT is an effective study aid, and that its safeguards against emotional dependence hold up under everyday teen use. "I think that like OpenAI's teen experience makes some really important safety commitments," Robbie Torney, head of AI and digital assessments at the Youth AI Safety Institute at Common Sense Media, told Mashable. "What we really need now is evidence that they work, and we need this evidence, including our testing and the testing of others, before anyone, including parents, puts real trust in ChatGPT for Teens." What should parents know about ChatGPT for Teens? Importantly, Torney said parents shouldn't assume their child is using ChatGPT for Teens. While OpenAI uses age estimation to place users in the age-appropriate version of ChatGPT, it can still mistake a user as older. Why you still need parental controls Parents can be sure their child is using the teen version by linking their ChatGPT account to their child's and implementing OpenAI's parental control features https://mashable.com/article/chatgpt-parental-controls . If a teen has previously lied about their age — and plenty do https://www.bbc.com/news/articles/cn4v52ezx17o — they may continue to access the adult version of ChatGPT. Parental controls allow caregivers to limit sensitive content and turn on Study Mode and Quiet Hours. The former restricts ChatGPT to learning content, while the latter shuts down ChatGPT altogether. Torney said caregivers must enact parental controls to receive notifications from OpenAI should their teen discuss suicide, self-harm, or acts of violence in immediate or dangerous ways. Even then, Torney said notifications may not reach parents immediately or quickly. Previous research conducted by the Youth AI Safety Institute at Common Sense Media designed to test notifications for self-harm and suicidal content found that such messages arrived a day or two after the exchange, or not at all. An OpenAI spokesperson said every notification is reviewed by a trained human before it's sent. The company strives to review notifications in under one hour. "You can view these features as backup, but they're not a replacement for your continued involvement in how your child is using the product, because we don't have confirmation or independent verification that they work," Torney said. Teens still may turn to ChatGPT instead of humans first Given the growing concern over teens replacing human relationships with AI chatbots, it's unsurprising that OpenAI launched ChatGPT for Teens as a product it claims supports real-world connections. "Supporting children’s cognitive development in the age of AI will require asking not only whether a product can be made safer for young people, but what role we want AI to play in their development in the first place," Camille Carlton, senior director of strategy and impact at the Center for Humane Technology, said in a statement. "Some of the most critical capacities children develop are not learned by interacting with a machine, but by spending time without one." Terms of Use https://www.ziffdavis.com/terms-of-use and Privacy Policy https://www.ziffdavis.com/ztg-privacy-policy . Lauren Jonas, head of youth and families at OpenAI, told Mashable that ChatGPT for Teens would redirect personal or emotional conversations back toward task-oriented discussions. "If they are not engaging in learning-focused content, the goal is to continue to bolster the age-appropriate experience and the safeguards," she said. However, Torney said Youth AI Safety Institute researchers conducting preliminary testing of ChatGPT for Teens were still able to have "personal support conversations" unrelated to learning. He said parents should watch for exchanges that feel more like companionship or mimicking of human connection, even if OpenAI believes ChatGPT for Teens shouldn't behave in that manner. Can ChatGPT really help with learning and studying? Though OpenAI said ChatGPT for Teens will stick to learning, the tool's ability to foster that is still up in the air, according to experts. The company's first attempt to turn ChatGPT into a homework helper was "fairly weak," explained Dr. Hamsa Bastani. Bastani is an associate professor of Operations, Information, and Decisions at the Wharton School, University of Pennsylvania, and a leading researcher into AI's potential learning gains. OpenAI launched Study Mode for ChatGPT https://mashable.com/article/chat-gpt-study-mode-review over a year ago, amid a major push to get the company's AI tools in the hands of students around the country. In Mashable's own testing, ChatGPT Study Mode still provided answers to homework questions when prompted and failed to generate appropriate study tools. Bastani reviewed Mashable's tests at the time, as well as our comparisons with competitors Google https://mashable.com/article/google-gemini-guided-learning-review and Claude https://mashable.com/article/anthropic-claude-learning-mode-review . She noted that many AI learning tools were repackaged Large Language Models LLMs with additional backend prompts to curb cheating. Bastani and her team had yet to find an effective AI tutor on the market as of last year. Bastani and her colleagues published their first, genuinely positive AI learning result https://papers.ssrn.com/sol3/papers.cfm?abstract id=6423358 just five months ago, long after Study Mode's launch. Bastani has not yet tested the upgraded version for teens, which now includes additional notifications to complete homework "responsibly." Jonas told Mashable that OpenAI is conducting randomized, controlled trials of Study Mode in a number of countries. Trusting teens to keep themselves on track The increasingly personalized, frictionless assistance provided by universal chatbots is counter to the principles of learning, some experts say. "This is especially important for teens, who may over-rely on general-purpose AI tools for tasks that should help them build their own skills," said Dylan Arena, chief data science and AI officer at education solutions company McGraw Hill. "It also matters in social contexts, where teens are still developing the skills and confidence to navigate real-world interactions. To be built for learning, chatbots must be dependable, repeatable, and safe." Arena agreed that an AI learning tool that attempts to form a relationship with the student is a red flag. The best educational tools — designed with limited use cases, safety guardrails, and measurable learning outcomes — often come with their own crash courses on how to actually use them effectively. Learners, in turn, have to be committed to sticking to those recommendations. That's a tall order for a teen just trying to get a bit of help with their schoolwork. "It seems like OpenAI is trying to put responsibility on teens to have agency and say, 'This is how I want to learn,'" said Torney. "Putting that responsibility on a still-developing brain that may be frustrated or stuck may lead to different outcomes that undermine learning." Bastani recently co-authored a study https://papers.ssrn.com/sol3/papers.cfm?abstract id=5604932 that explored this concept of "self regulation," which has become a key component in the . Researchers compared learning outcomes from chess academy students who received automatic, pre-programmed hints from AI versus students who could request help from AI at any time. https://mashable.com/article/ai-k-12-classrooms-ban-debate AI in education debate "Poor self-regulation caused students to increasingly over-rely on AI assistance even when they expressed that they were worried it would harm learning," she said of the findings. A recent Common Sense Media report on teen AI use https://www.commonsensemedia.org/research/teens-in-the-ai-era-schoolwork-and-skills-that-matter for schoolwork found another noteworthy pattern: A majority of students still lean on AI for answers, even those who use chatbots in appropriate, recommended ways. Torney said a number of questions remain unanswered about learning and ChatGPT for Teens, including the efficacy of its step-by-step guidance and responsible homework reminders. "How much does this prevent kids from taking shortcuts when we know that those shortcuts undermine learning?" Torney asked. Disclosure: Ziff Davis, Mashable’s parent company, in April 2025 filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems. Topics Social Good https://mashable.com/category/social-good Family & Parenting https://mashable.com/category/family-parenting ChatGPT https://mashable.com/category/chatgpt Rebecca Ruiz is a Senior Reporter at Mashable. She frequently covers mental health, digital culture, and technology. Her areas of expertise include suicide prevention, screen use and mental health, parenting, youth well-being, and meditation and mindfulness. Rebecca's experience prior to Mashable includes working as a staff writer, reporter, and editor at NBC News Digital and as a staff writer at Forbes. Rebecca has a B.A. from Sarah Lawrence College and a masters degree from U.C. Berkeley's Graduate School of Journalism. Chase joined Mashable's Social Good team in 2020, covering online stories about digital activism, climate justice, accessibility, and media representation. Her work also captures how these conversations manifest in politics, popular culture, and fandom. Sometimes she's very funny.