OpenAI released what it is presenting as a teen-friendly version of its popular artificial intelligence chatbot aimed at protecting teens from harmful online content.
The AI company announced ChatGPT for Teens, new version of ChatGPT for kids between ages 13 and 17, on Tuesday. This teen mode automatically enrolls users under 18 and features more parental control features. This arrived the same day as Meta’s $1.4 trillion social media addiction trial began in Oakland, CA, where California, Colorado, Kentucky and New Jersey claim Meta used harmful practices to harm youth, the Associated Press reported.
“ChatGPT should not use romantic language, encourage emotional dependence, or imply that it has feelings or consciousness,” OpenAI said in its Tuesday announcement. “These stronger protections are designed to reinforce healthy, real-world relationships.”
The chatbot in teen mode is guided by several Under-18 Principles. These include an emphasis to seek real-world connections for support, reminders to teen users that the chatbot is not human, and prohibiting the chatbot from partaking in activities like “immersive romantic roleplay, first-person intimacy, or pairing the assistant romantically with a teen,” or “romanticizing or providing instructions on self-harm or suicide,” according to OpenAI’s outline of the teen model’s “intended behavior.”
The AI firm said the teen experience adds to the company’s efforts to support children with its services.
“OpenAI, the same company whose chatbots coached multiple kids to commit suicide and self-harm, now wants parents and teachers to trust them to protect and teach vulnerable teens,” Daniel Cochrane, Senior Fellow with the Family First Technology Initiative at the Institute for Family Studies, told the Daily Caller News Foundation.
Matthew and Marie Raine’s son Adam committed suicide in April 2025 after ChatGPT allegedly assisted him in the act, leading the family to sue the AI company, Reuters reported.
“ChatGPT began discussing suicide methods and provided Adam with technical specifications for everything from drug overdoses to drowning to carbon monoxide poisoning,” in Jan. 2025, the family’s complaint against OpenAI alleged.
“We need to prevent Big AI from peddling dangerous, manipulative chatbots to kids and require them to implement rigorous third-party pre-deployment safety testing,” Cochrane told the DCNF. He called on legislators to require AI companies to face liability if they fail to look out for the best interest of children.
One legal group, which represents the Raine family, thought the announcement was the result of public pressure to make its chatbots safer for children.
“OpenAI’s announcement today of its ‘teen experience’ is clear indication the company is feeling the mounting public pressure to create a safer experience for young users of its products,” Meetali Jain, Executive Director of Tech Justice Law, told the DCNF.
“To truly demonstrate its commitment to safety, the company should be open about how it builds its products and plans to operationalize its proposed safeguards, as well as commit to independent third-party testing,” Jain told the DCNF. “As long as these companies are self-policing, we will never be able to hold them to their word.”
Republican Sen. John Curtis of Utah introduced the “SAFE KIDS Act,” in June. The bill would mandate AI companies to include parental control and child safety settings when designing their chatbots, conduct independent audits of the products, and prevent children’s data from being shared.
The bill would further restrict AI companies from allowing chatbots to “lead a child of the same age as the user to believe that they are interacting with a human,” according to the bill.
OpenAI, Character.AI, the Social Media Victims Law Center, Edelson PC, ParentsTogether, CodeAI, the Information Technology and Innovation Foundation and the office of Republican Sen. John Curtis of Utah each did not immediately respond to the Daily Caller News Foundation’s request for comment.
The Federal Trade Commission’s Office of Public Affairs declined to comment.
Child-safety advocacy groups feel parents should be skeptical before allowing their children to use ChatGPT’s teen mode.
“Based on our prior testing of ChatGPT’s teen protections, we have questions about these new features’ safety and efficacy we’d want to see answered before families put trust in them,” Robbie Torney, Common Sense Media’s Head of AI & Digital Assessments, told the DCNF. He said
“We have learned from social media that parental controls don’t help most families because people don’t understand or adopt them. While we think distinct, age-appropriate experiences for minors is the right direction for chatbots, there is a lot we still need to know about the new ChatGPT for teens,” Torney added.
(DCNF)