OTTAWA — OpenAI is rolling out a new version of ChatGPT specifically for teenagers, adding stronger restrictions around violence, self-harm and other sensitive content months after the company came under scrutiny in Canada over its handling of the Tumbler Ridge school shooter.The company announced ChatGPT for Teens on Tuesday, saying users between 13 and 17 will receive an experience with additional safety protections and controls for parents.The new protections include prompts telling teens to take a break after 90 minutes of use within a three-hour period, along with "Quiet Hours" that can prevent ChatGPT from being used at certain times. Parents or teens can also establish "Study Hours," which automatically turns on the platform's learning-focused study mode. OpenAI said the new experience applies "age-appropriate safeguards designed to reduce exposure to content that may be harmful or developmentally inappropriate," while still allowing teenagers to "learn, create, and explore." Users who say they are under 18 when signing up will automatically receive the additional protections. OpenAI is also using artificial intelligence to estimate a user's age and says accounts believed to belong to minors can be placed under the restrictions regardless of the birthdate entered on the account.The company said it has strengthened protections against graphic content, sexual or romantic interactions and violent roleplay. It has also made changes intended to prevent ChatGPT from expressing personal feelings toward teenage users and will warn teens to think twice before uploading private or sensitive images. Parents can already receive alerts when ChatGPT detects "a moment of acute distress" involving their child. OpenAI said Tuesday it plans to expand those notifications in the coming weeks to include concerns involving eating disorders. .The new teen service was not announced as a direct response to the Feb. 10 mass shooting in Tumbler Ridge, B.C., but several of the safeguards address areas that came under scrutiny in Canada following the attack, including violent content and how potentially dangerous behaviour is handled.OpenAI revealed following the shooting that it had banned a ChatGPT account belonging to 18-year-old Jesse Van Rootselaar in June 2025 after detecting what the company described as "misuses of our models in furtherance of violent activities."The company considered contacting police but ultimately decided the activity did not meet its threshold for referral because it could not identify credible or imminent planning. Van Rootselaar was a transgender man, RCMP officials said the 18-year-old had transitioned about six years before the shooting, at approximately age 12. Van Rootselar, who was born male, later killed his mother and half-brother at their home before going to Tumbler Ridge Secondary School, where five students and an educational assistant were killed. Van Rootselaar then died from a self-inflicted gunshot wound.The revelation that OpenAI knew about the violent activity months earlier prompted Artificial Intelligence Minister Evan Solomon to summon the company's senior safety officials to Ottawa in February."We made it clear that Canadians expect credible warning signs of serious violence to be escalated in a timely and responsible way," Solomon said following the meeting."Internal review alone is not sufficient when public safety is at stake." Solomon also expressed frustration that OpenAI did not arrive at the meeting with new measures to present to the government."We expressed our disappointment that no substantial new safety measures were presented at this time," he said. "OpenAI indicated they will return shortly with more concrete proposals tailored to the Canadian context." OpenAI responded at the time by saying it had heard the government's concerns."The ministers underscored that Canadians expect continued concrete action and we heard that message loud and clear," the company said..OpenAI committed to follow up with additional steps and said it had already strengthened safeguards and changed its law-enforcement referral protocols for cases involving violent activity. The company later faced lawsuits from seven families connected to the Tumbler Ridge shooting alleging ChatGPT was complicit in the deaths or injuries of their children because Van Rootselaar had extensive conversations with the chatbot involving gun violence. OpenAI has disputed allegations against the company and said it strengthened its guardrails. CEO Sam Altman also publicly apologized to the Tumbler Ridge community in April over the company's failure to alert authorities about Van Rootselaar before the attack. Tuesday's launch of ChatGPT for Teens comes as the physical site of the shooting is being torn down.Crews began removing the exterior of Tumbler Ridge Secondary School this week as part of a staged demolition of the building. School District 59 has said the structure will take several weeks to remove before the site is restored to grass. "We recognize that seeing the building change will be significant for many in the community," School District 59 superintendent Christy Fennell said ahead of the demolition. A replacement school is planned at a different location, while students will use temporary facilities during construction.
