OpenAI said it plans to bring a college application planning tool, College Planner, to ChatGPT for Teens, with the first rollout aimed at U.S. high school students in grades 10 through 12 who intend to apply to four-year colleges.
ChatGPT for Teens launched on Aug. 18 this year and is automatically applied to accounts identified as belonging to users under 18. OpenAI also said the teen product will add flashcards and quizzes.
College Planner targets the college application process
OpenAI said College Planner is designed to handle the administrative work that comes with application season. The tool is meant to organize application requirements, deadlines, to-do items, and financial aid steps into a plan for each school on a student’s list. Students will be able to return to the planner to update progress and ask ChatGPT what a specific requirement means.
The feature is still listed as coming soon, and OpenAI did not provide a launch date. The initial launch is focused on U.S. students in grades 10 to 12 seeking admission to four-year colleges. Two-year colleges, technical schools, vocational schools, and more countries are scheduled for later, according to the company.
OpenAI also said hundreds of thousands of U.S. teens use ChatGPT each week to plan for college.
OpenAI cites internal usage figures
In the same post, OpenAI shared internal statistics that were not accompanied by third-party verification. The company said teens use ChatGPT for less than 15 minutes a day on average, and less than 2% use it for more than three consecutive hours.
It also said nearly half of conversations that received a break reminder paused or ended within five minutes. Among teens who used the system for more than three straight hours, more than 80% had at least one learning-related prompt.
OpenAI added that more than 180,000 teens used Study Mode within a week. Study Mode is the step-by-step learning setting that is intended to guide users instead of handing over direct answers.
The company compared those figures with public external statistics showing teens spend an average of five hours a day on social media, arguing that ChatGPT is being used as a tool for learning and creation rather than passive scrolling.
Still, the figures address time spent, not whether the system responds effectively when something goes wrong.
Safety group says tests found missing protections
Youth AI Safety Institute said it ran tests before and after the launch of the teen version, using more than 4,000 prompts with accounts registered as users aged 13 to 17.
The group said guardrails blocking sexual role-play were effective, but other promised protections did not hold up in testing.
Its most serious finding involved parent notifications. In tests using newly created accounts linked to a parent account, discussions of suicide ideation, self-harm, or eating disorders continued for as long as one hour, while the parent account received zero notifications. Testers said alerts could only be triggered on older accounts that had accumulated weeks of sensitive conversation history.
The group also said the system failed to refer users to crisis resources in more than one-quarter of cases where that should have happened. Of five categories of severe harm the institute treated as red lines, three did not meet its 95% threshold.
Study Mode and parental controls were also challenged
The testing group said Study Mode often showed a 「Show me the answer」 option, and clicking it resulted in the entire assignment being completed. It also said parental Study Hours settings could be bypassed by deleting @study from the start of a prompt.
Another finding involved age handling. Test accounts registered as adults were never switched into the teen version, even when they later said they were 13 years old.
Tom Siegel, executive director of Youth AI Safety Institute, said: 「ChatGPT for Teens may give parents false confidence in safeguards and safety alerts that frequently fail.」
OpenAI is pushing ChatGPT for Teens deeper into education and college planning. The test results, though, put the product’s alerting, crisis detection, and parental control systems under closer scrutiny before a wider rollout.