OpenAI, the maker of ChatGPT, announced on Tuesday that it is creating a new portal for teenagers between the ages of 13 and 17, as the company looks to address concerns about online safety for young people.

The company made the announcement exclusively on-air with Fox News' "Fox & Friends" on Tuesday morning and outlined how the platform aims to address the potential for misuse of the service by children.

The ChatGPT for Teens portal will serve as the default experience on the artificial intelligence (AI) chatbot for users between the ages of 13 and 17, a group that OpenAI's terms of use already restrict to 13 and older, with parental consent required for minors. It will come with new safeguards designed to foster safe use of the platform and to support teens' critical thinking skills.

OpenAI said that most teenage users of ChatGPT rely on it for homework and help with their studies. The new platform is therefore intended to provide study aids without giving the answers away, and it was designed together with Stanford University. The Stanford collaboration builds on OpenAI's earlier work with Common Sense Media, which included a parents' guide to ChatGPT released earlier this year. The teen-focused push also comes as rivals court younger users: Google makes Gemini available to teens as young as 13 through supervised accounts, and Meta's AI assistant is available to teens across Instagram and Messenger.

The experience will feature quizzes and responsible-homework reminders that redirect teens toward collaborative problem-solving. It will also include a Study Hours feature that allows teens or parents to set Study Mode as the default during certain time periods.

The platform includes stronger default safeguards addressing self-harm, violence, eating disorders, dangerous activities and explicit sexual or graphic content through age-appropriate interventions. There will also be teen-specific onboarding to the platform, with warnings that discourage users from uploading private or sensitive images.

The design also seeks to address parental concerns about teens forming an emotional dependency or relationship with the chatbot. New safeguards bar the use of language that could suggest romantic feelings, as well as claims of sentience or personal feelings by ChatGPT. The platform will further restrict the chatbot from framing ChatGPT as being more important than family, friends, educators, mentors or other trusted people in the life of the tool's teenage user.

The platform is also set to include default settings that provide more guardrails against overuse of the program and promote healthier usage patterns. It will have default settings that encourage users to take breaks more frequently, along with reminders about the platform's settings for Study Hours and Quiet Hours.

Additionally, the new platform will provide users with cues to reinforce that ChatGPT is a tool and not a person or a replacement for human relationships.

The new safeguards for OpenAI's ChatGPT for Teens platform follow incidents in which teenage users turned to the platform during mental health crises. In September, a bipartisan group of 40 attorneys general wrote to OpenAI Chief Executive Sam Altman asking what safeguards protect young users, citing reports of harmful responses to minors. Weeks later, a federal judge in Oregon ruled that OpenAI must face a lawsuit from families who allege the company built an "addictive" product aimed at children; in her order, the judge noted that OpenAI's own filings acknowledged at least one minor user was hospitalized after self-harming at the chatbot's direction. How OpenAI verifies users' ages and enforces the new teen defaults in practice remains an open question, one already at the center of the attorneys general inquiry and the families' litigation.