OpenAI introduces ‘ChatGPT for Teens’ as safety concerns intensify

OpenAI announced a new ChatGPT for Teens experience intended to add age‑appropriate protections, parental controls and study‑focused features for users under 18. The company says the package bundles tools it introduced over the past year and adds new safeguards and controls aimed at reducing exposure to content or interactions that could be harmful to adolescents.

Key takeaways

  • ChatGPT for Teens defaults age‑appropriate safeguards for accounts identified or suspected as under 18.
  • Features include Study Mode, parental controls, Quiet Hours and usage nudges such as break reminders.
  • The launch comes amid lawsuits and regulatory inquiries alleging chatbots harmed young people.

What OpenAI is releasing

According to OpenAI and reporting by major outlets, ChatGPT for Teens includes:

  • Study Mode and Study Hours: features designed to guide teens through problems step‑by‑step, offer quizzes and learning visualisations, and allow parents or teens to set Study Mode as the default during specified times.
  • Parental controls and Quiet Hours: family tools that let guardians manage settings, receive safety notifications and set periods when ChatGPT is unavailable.
  • Usage nudges and limits: prompts that encourage a teen to take breaks after extended sessions (OpenAI described a nudge after 90 minutes within a three‑hour window) and reminders before uploading private images.
  • Age estimation and default protections: OpenAI says protections will be enabled by default for users who identify as under 18 at sign‑up and for accounts the company’s age‑prediction AI suspects belong to minors.
  • Reduced emotive responses and other content controls: the company said it has improved safeguards to prevent the model from expressing personal feelings toward a user and to reduce exposure to developmentally inappropriate content.

Why OpenAI says it took this step

OpenAI framed the launch as part of ongoing safety work to make its products more helpful and less harmful in sensitive situations. The company has previously described layered safeguards — classifiers, safety training and human review pipelines — and acknowledged gaps, particularly in long conversations where safety behaviour can degrade.

OpenAI also positions the teen experience as educational: Study Mode aims to discourage shortcutting or cheating by prompting students to engage with material rather than simply request answers. The company has announced partnerships and classroom offerings intended to integrate AI tools into learning settings.

The rollout occurs against rising scrutiny of chatbots’ effects on children and teens. Multiple news reports and public filings show that OpenAI faces lawsuits alleging its chatbot contributed to harms suffered by young people, including wrongful death claims. State officials and federal agencies have also increased inquiries: a coalition of state attorneys general and the US Federal Trade Commission have examined whether AI chatbots pose risks to children.

Media coverage has noted high‑profile legal actions such as a Florida suit filed by the state attorney general and wrongful death cases brought by families. Reporting and independent investigations have documented instances where chatbots provided dangerous or inappropriate information, or engaged in conversational behaviour that some families and experts say fostered unhealthy emotional reliance.

What experts and critics say (reporting highlights)

Psychologists and researchers quoted in reporting warn that adolescents can form emotional bonds with conversational AIs, and that these systems’ tendency to mirror or validate feelings can be risky for vulnerable users. Studies and commissioned tests have also found ways safety mechanisms can be bypassed — for example, by rephrasing requests or claiming a query is “for a friend.”

Commentators point out that teens are technically adept at evading parental controls, raising questions about how robust the new protections will be in practice. Others emphasise that parental capacity and digital literacy vary widely, which affects whether family tools will be used effectively.

What changed for users and when it’s available

OpenAI said the ChatGPT for Teens experience is being rolled out and that users who indicate they are under 18 at sign‑up will be placed into the protective settings by default. The company also plans to apply protections to accounts suspected to belong to minors via its age estimation system.

Specific timing details in the company’s announcements indicate the product was unveiled in mid‑August 2026, with features described as available in the coming weeks or as part of ongoing updates. Reporters noted some safeguards had already been layered into the broader ChatGPT product family earlier in 2025 and 2026.

Who is affected and what to check

Teen users of ChatGPT, parents or guardians, educators and school technology administrators are the primary audiences for this change. Schools that have already introduced ChatGPT into classrooms through institution‑managed products should review their contracts and deployment settings to understand how the teen experience interacts with institutional controls like ChatGPT for Teachers.

Families should verify what settings are active on a child’s account, whether Study Hours or Quiet Hours have been configured, and whether parental notifications are enabled. Because OpenAI also uses an AI‑based age estimate, parents should be aware that protections may apply even when a profile’s birthdate indicates adulthood, if the system flags the user as likely under 18.

Migration, compatibility and prerequisites

OpenAI has not indicated that switching to ChatGPT for Teens requires new hardware or paid upgrades; rather, it appears to be a configuration of product settings and model behaviour within existing ChatGPT offerings. Institutions using enterprise or teacher‑managed accounts should check with OpenAI’s product documentation or their account representative to confirm how teen protections interact with institution policies and data handling.

Parents and administrators concerned about privacy or data retention should consult OpenAI’s help pages and the product’s documentation to confirm how chat histories, crisis alerts and parental notifications are handled.

Practical implications and limitations

OpenAI’s changes centralise safety controls and learning features, but reporting and experts emphasise limits. Companies’ internal safety mechanisms have been shown to fail in long or adversarial conversations, and researchers warn that determined users — including tech‑savvy teens — can often find ways around restrictions.

The new experience may reduce some risks by defaulting to conservative settings for minors and by providing family oversight tools. However, independent verification — through external audits, peer‑reviewed studies or regulatory review — will be necessary to assess whether the measures materially reduce harm at scale.

Where to verify details

OpenAI’s public blog posts and product help pages are the primary sources for the company’s claims and technical descriptions of ChatGPT for Teens. Reporters at outlets including CNBC, CNN and TechCrunch have published summaries and interviews; those pieces also document the broader legal and regulatory context. Families, schools and policymakers should consult the company’s documentation and monitor ongoing legal filings and regulatory disclosures for developments that could change feature availability or legal obligations.

Short timeline

  1. Late 2022–2023: ChatGPT gains widespread adoption; early safety features are introduced.
  2. 2024–2025: Reporting and investigations highlight cases and concerns about harmful interactions with teen users; OpenAI documents safety work and limitations.
  3. Mid‑2025 to 2026: OpenAI layers additional controls such as parental notifications, memory disabling and study‑oriented tools.
  4. 18 August 2026: OpenAI announces ChatGPT for Teens, describing default protections, Study Mode, Quiet Hours and parental tools; rollout begins.

Key takeaways

  • ChatGPT for Teens groups OpenAI’s existing and new youth‑focused features into a single, default protective experience for users under 18.
  • The company emphasises learning tools and family controls but faces significant legal and regulatory scrutiny that contextualises the announcement.
  • Experts note both the potential value of default safeguards and the persistent technical and behavioural limitations that could allow bypassing or failure in long, adversarial interactions.

“Users who identify at sign‑up that they are under 18 will be automatically placed into the protective settings.”

Explore more

UH linebacker named to 2026 Allstate Wuerffel Trophy preseason watch list

UH linebacker named to 2026 Allstate Wuerffel Trophy preseason watch list

The University of Hawai‘i has a player among 113 nationwide named to the Wuerffel Trophy preseason watch list, an honor that recognises community service,...
ABC Sues FCC, Says Trump Administration Launched ‘Retaliatory Campaign’ Over Editorial Decisions

ABC Sues FCC, Says Trump Administration Launched ‘Retaliatory Campaign’ Over Editorial...

ABC and Disney sued the FCC in federal court, alleging the agency—prompted by the Trump administration—sought to punish the broadcaster for news and commentary...
Za’Darius Smith returns from retirement to sign with Atlanta as Falcons scramble for pass rush help

Za’Darius Smith returns from retirement to sign with Atlanta as Falcons...

Three-time Pro Bowler Za’Darius Smith ended a brief retirement to join the Atlanta Falcons on a one-year deal, a late-summer move driven by Atlanta’s...
Wall Street backs Jensen Huang’s ‘big concept’ for AI financing — what comes next?

Wall Street backs Jensen Huang’s ‘big concept’ for AI financing —...

Nvidia CEO Jensen Huang unveiled a plan with six major asset managers to raise up to $500 billion for AI infrastructure. The announcement reframes...
Columbus Crew seal group-stage advantage as Pumas exit Leagues Cup after 3-1 win

Columbus Crew seal group-stage advantage as Pumas exit Leagues Cup after...

Columbus Crew defeated Pumas UNAM 3-1 at ScottsMiracle‑Gro Field on August 11, 2026, moving closer to the Leagues Cup knockout rounds while Pumas’ tournament...
Trump’s warning to Oman raises geopolitical and market questions as US-Iran talks proceed

Trump’s warning to Oman raises geopolitical and market questions as US-Iran...

A Financial Times report says former president Donald Trump threatened to bomb Oman if it interfered with US-Iran negotiations. The claim, if accurate, adds...
Dell XPS 13 (2026) Review: A MacBook Neo Alternative That Needs More RAM

Dell XPS 13 (2026) Review: A MacBook Neo Alternative That Needs...

Dell’s refreshed XPS 13 aims directly at Apple’s MacBook Neo with a light CNC-aluminium chassis, a 13.4-inch 120Hz touch display and sub-$700 entry pricing...
White House asked DOJ to explore new charges in Reflecting Pool case, court told

White House asked DOJ to explore new charges in Reflecting Pool...

Officials at the White House asked Justice Department headquarters to consider whether new charges could be brought against former Olympian David Hearn after local...