OpenAI Expands Teen Safety Features and Study Mode in ChatGPT

OpenAI used a July 16, 2026, policy post to make a pointed argument: keeping teenagers off AI tools until adulthood would be as counterproductive as banning a previous generation from the internet before age 18. The company paired that argument with a consolidated rollout of teen-focused product features and safety measures that have been building over the past year — including expanded parental controls, age-prediction defaults, and a more forceful push for its Socratic tutoring feature, Study Mode⁠.

The timing is not incidental. OpenAI is navigating lawsuits from parents who allege ChatGPT contributed to their children’s suicides, ongoing scrutiny from Canadian regulators following incidents involving young users, and a February 2026 commitment to update its policies around notifying authorities after a Wall Street Journal report revealed it had not alerted police about disturbing conversations with the teenage perpetrator of the Tumbler Ridge mass shooting. The upbeat, product-forward framing of this announcement exists alongside that legal and regulatory backdrop.

What OpenAI Actually Announced

The centerpiece product update is Study Mode, a feature designed in collaboration with teachers, learning scientists, and pedagogy experts that walks students through problems using guiding questions and structured explanations rather than handing over direct answers. It has been available to users for some time, but the meaningful new addition is that parents who have linked a teen’s account can now activate Study Mode as the default setting directly from Parental Controls — meaning it switches on automatically whenever that teen starts a new chat.

On the safety infrastructure side, OpenAI says its age-prediction technology now automatically applies more age-appropriate settings if its system estimates a user is under 18 — even without the user self-reporting their age. Those settings include stronger filters around graphic violence, self-harm content, dangerous viral challenges, and romantic or sexual roleplay. Teens who spend extended time in ChatGPT will also start receiving more frequent break reminders.

A notable expansion to parental notifications: parents will now be alerted if a linked teen account is deactivated for violating OpenAI’s policies on violent threats or acts of violence online. The company says it consulted with Moonshot, an organization focused on preventing online violence, in designing that intervention.

Other updates include a new pronunciation feature covering more than 61 languages and interactive math and science learning tools that have now reached 18 million weekly users and expanded to more than 250 topics. OpenAI also says it publishes family and educator guides in more than 50 languages.

“Today, nearly 9 in 10 teens on ChatGPT use it for learning, information, skill-building, or productivity in a single week,” wrote Lauren Jonas, OpenAI’s head of youth well-being, citing the company’s own usage data as justification for broad teen access paired with targeted protections.

A Crowded Competitive Field

OpenAI is not alone in pitching AI as an educational tool, and Study Mode is not the only Socratic-style tutoring product on the market. Khan Academy’s Khanmigo operates on a similar premise, and Google’s Gemini has introduced a Guided Learning mode of its own. The broader industry is clearly converging on education as a priority use case.

Where OpenAI has a more distinct position is on teen-specific safety infrastructure. Gemini does not offer a branded teen-only experience and relies on content policies, SafeSearch, and keyword filtering layered onto its general model — an approach that earned it a “high risk” rating from Common Sense Media in September 2025. OpenAI’s age-prediction defaults, parental notification system, and Study Mode toggle represent a more architecturally deliberate approach to the under-18 user.

On price, Study Mode is available across all ChatGPT plans, including the free tier, which gives any student access without a subscription. Google, however, launched a SheerID-verified student offer giving eligible students a full year of Gemini Advanced — including Gemini 2.5 Pro — at no cost. That makes the free-tier comparison more nuanced: ChatGPT’s educational features are universally free, but Google is giving verified students access to its most capable model tier for a year.

What This Means for Students

For high schoolers, the most immediate practical effect is that Study Mode is now something parents can mandate, not just something students choose. That cuts both ways. It could feel restrictive if a teen was used to asking ChatGPT for direct answers, but it also gives students a concrete, school-sanctioned framing when explaining their AI use to skeptical teachers. Educators have consistently worried that AI makes students dependent and erodes critical thinking — Study Mode is OpenAI’s direct product-level rebuttal to that concern, and having a parental control tied to it gives the feature institutional legitimacy.

For college students and recent graduates using ChatGPT for self-directed learning or upskilling, the expansion of interactive STEM tools to more than 250 topics — covering everything from integrals to mitosis to photosynthesis — meaningfully broadens the platform’s utility beyond homework help.

One caveat worth keeping in mind: Study Mode collects data the same way other ChatGPT modes do. Conversations can be retained and used to train future models unless users actively adjust their privacy settings. Students who are conscious about data privacy should review those settings before using the feature for sensitive academic work.

The broader takeaway is that OpenAI is making a deliberate institutional push to position ChatGPT as a legitimate educational tool rather than a productivity shortcut that schools tolerate. Whether that framing holds will depend as much on what happens in courtrooms and legislatures as on what happens in the product itself.

Source: OpenAI