Most advertised child safety features on Instagram, Snapchat, TikTok and YouTube do not work as promised or are too difficult for young users to find, according to a new independent audit that raises new questions about how technology companies protect children online.
The report, titled “Broken, Buried, or Missing: Anatomies of Failure (and Success) of Social Media Child Safety Features,” was published June 29 by the Cybersafety Research Center, a joint research initiative of New York University and Northeastern University. It was produced with support from Heat Initiative, a nonprofit that campaigns against online child sexual exploitation. The center describes it as the first independent audit of its kind covering the four platforms.
Researchers tested 86 advertised child safety features across the four platforms. Only 35 both worked as described and were accessible enough for a young user to realistically find and use, the report found. The remaining 51 were classified as missing, broken, or buried, meaning they could not be triggered, did not function as claimed, or were hidden too deep in settings to be practical for a child to use.
Snapchat had the highest failure rate, with 8 of its 11 tested features failing, or about 73 percent. Instagram followed with 19 of 29 features failing, a 66 percent rate. YouTube had 12 of 22 features fail, or 55 percent, and TikTok had 12 of 24 fail, a 50 percent rate, the report said.
Researchers said they evaluated each feature on two measures, whether it functioned the way the company described, and whether a child could reasonably find and activate it. Tools that required several difficult steps, were buried deep in settings menus, or failed to activate during testing were not counted as successful.
Among the more serious concerns were failures involving search restrictions meant to keep children from finding harmful material. All four platforms advertise tools that intercept dangerous search terms and redirect young users toward support resources. Researchers said the protections on Instagram, Snapchat and TikTok could be defeated simply by misspelling a search term or leaving it incomplete, and that Instagram’s and TikTok’s own built-in search suggestions sometimes offered alternative wording that led directly around the restrictions. On YouTube, researchers said a similar restriction could be bypassed by clicking through a warning screen.
The four companies disputed parts of the findings when contacted by CNN. A TikTok spokesperson pointed to more than 50 preset safety settings automatically applied to teen accounts, along with its Family Pairing tool for parents. Snapchat said many of the findings reflected deliberate attempts to bypass safeguards rather than how teenagers typically use the app. Meta and YouTube said their time-limit and break-reminder tools work as intended, and said parental controls on screen time should not be judged by the same standard as general in-app reminders.
The findings carry particular weight for parents who rely on platform settings to judge whether a child is ready to use social media. The report does not argue that safety tools are useless. Researchers pointed to some features, including default private accounts for teens on Instagram and a restricted, view-only mode for younger TikTok users, as examples of designs that worked as intended. But it says these tools, taken as a whole, should not be treated as a complete protection system. 3A1Z Stories, a family and parenting platform has also covered the findings and said that checking privacy, messaging and screen-time settings directly inside each app, along with talking with children about their online experience, remains worthwhile regardless of how any single feature performs.
Governments around the world have increased pressure on technology companies over child safety in recent months. The European Union’s Digital Services Act requires platforms accessible to minors to take measures to protect them and bans targeted advertising to children based on profiling.
Australia has enforced a ban on social media accounts for children under 16 since December 2025. In the United Kingdom, the government announced June 15 that it would go further than its existing Online Safety Act, barring children under 16 from platforms including Instagram, Snapchat, TikTok and YouTube in a model similar to Australia’s, with the new restrictions expected to take effect in spring 2027.
For parents, educators and policymakers, the findings add to a broader debate over whether voluntary safety tools are enough on their own, or whether social media platforms need independent testing and stronger accountability before advertising their products as safe for children.
————————————————
