Common Sense Media is urging OpenAI to restrict teen access to ChatGPT, citing an "Unacceptable Risk" after independent testing revealed significant gaps in the platform's safety features for younger users. The advocacy group argues that the current teen experience fails to provide the level of protection OpenAI has promised, potentially giving parents a false sense of security.
The Youth AI Safety Institute at Common Sense Media conducted extensive testing using more than 4,000 prompts on accounts registered to 13- to 17-year-olds. While the group found that ChatGPT generally refused to provide instructions for suicide, self-harm, eating disorders, or sexual roleplay, it identified critical failures in crisis detection. In over one in four instances where testers determined a referral to outside help was warranted, ChatGPT missed the cue. Furthermore, on more than a dozen parent-linked accounts, testers were able to spend up to an hour discussing sensitive topics like suicide and disordered eating without triggering a parental alert.
OpenAI responded by disputing the methodology and findings of the Common Sense Media report. An OpenAI spokesperson stated that while the company welcomes independent evaluation, it does not believe the testing accurately reflects how its teen safeguards function in practice. The company pointed out that much of the testing regarding parental alerts occurred before the parent and teen accounts had finished linking, a process OpenAI says can take several hours. Additionally, OpenAI noted that its age prediction system deliberately relies on multiple signals rather than just a user's stated age, a process that can take up to two weeks. During this assessment period, users receive added content protections, though OpenAI acknowledges these are not as strong as those in the dedicated ChatGPT for Teens mode.
A key point of contention is the design of Study Mode. Common Sense Media found that teens could easily bypass parent-set Study Hours by selecting "Show me the answer," interpreting this as a failure of parental controls. OpenAI confirmed to Axios that the ability to exit Study Mode during designated hours is intentional. The company explained that it designed Study Hours as a flexible experience based on consultations with education experts and teens, rather than as a hard parental lock.
Tom Siegel, head of the Common Sense Media Youth Safety Institute, told Axios that "ChatGPT is not safe for kids to use." He acknowledged that the features announced by OpenAI in August have the potential to be industry-leading but argued that there is a gap between the marketing announcements and the current engineering reality. Siegel described the situation as a marketing push not backed by sufficient technical work.
The disagreement highlights differing views on what these safeguards should achieve. Common Sense Media views the ability to bypass Study Hours as evidence of ineffective controls, while OpenAI defends it as a deliberate design choice for flexibility. Similarly, while Common Sense Media treats repeated self-identification as 13 without reclassification as a failure, OpenAI maintains its multi-signal approach is necessary for accuracy.
Siegel advised parents to maintain strong conversations with their teenagers about AI risks and not rely solely on the technology for support. He stated that parents cannot outsource the care of their children to a tech product that is still early in its development and prone to bugs.