A new report released Monday by the Cybersafety Research Center — a joint initiative of New York University and Northeastern University — found that most of the child-safety tools promoted by major social media platforms failed to work as advertised, raising new questions about online protections for young users.
The research team, led by Laura Edelson, an assistant professor of computer science at Northeastern University, tested 86 youth safety features across TikTok, Instagram, Snapchat, and YouTube. Only 35 worked as promised. The remaining 51 either failed outright, were difficult to access, or could not be triggered despite following the companies’ published instructions. The report is titled “Broken, Buried, Missing.”
Researchers created both teen and adult test accounts on each platform, evaluating whether the safety features actually functioned as described and whether young users could realistically find and use them.
The findings were stark.
Nine safety features were classified as “missing” because researchers could not activate them at all. Thirty-four were labeled “broken,” meaning they failed to work properly or could easily be bypassed. Twelve of those were both broken and deeply buried within settings menus, while another eight technically worked but were hidden where most teenagers were unlikely to find them.
Among the most troubling findings involved TikTok. Researchers said the platform is designed to prevent minors from searching for content related to eating disorders and self-harm, yet its recommendation system instead suggested pro-anorexia search terms and self-harm phrases to a teen account.
Edelson emphasized those recommendations came directly from TikTok’s own algorithm rather than from search terms entered by the research team.
The study also found every platform struggled to moderate abusive behavior.
Safety tools designed to discourage bullying or harmful interactions failed across all four services. An Instagram feature intended to prompt users to reconsider before posting abusive comments never activated when researchers used a test account to harass another user. Many moderation systems relied heavily on blocked-word lists, allowing users to bypass protections simply by misspelling offensive words.
Features intended to limit excessive screen time also performed poorly, functioning successfully only about one-third of the time during testing.
The findings carry significant business implications beyond child safety.
For years, major social media companies have cited expanding safety features as evidence they can effectively regulate themselves without additional government intervention. The report arrives as lawmakers continue debating new online child-protection legislation, with technology executives expected to face renewed congressional scrutiny later this year.
If regulators conclude existing safeguards are ineffective, technology companies could face stricter compliance requirements and additional regulatory costs.
The platforms disputed many of the findings.
A YouTube spokesperson said the company has spent more than a decade developing parental controls and cited survey data showing most parents who use its supervised experiences report greater confidence in their children’s online activity.
A TikTok U.S. spokesperson said teen accounts include more than 50 safety settings enabled by default, maintained that the company’s internal testing confirms those features function properly, and offered to demonstrate them to the researchers.
The report also highlighted several positive examples.
Instagram automatically makes new teen accounts private, while TikTok’s experience for users under age 13 restricts commenting and direct messaging.
Edelson said those approaches point toward a simpler solution: make the safest settings the default rather than expecting parents and children to locate and activate them manually.
For parents, the report serves as a reminder that enabling a safety feature does not necessarily guarantee protection. For technology companies, it adds to growing pressure to demonstrate not only that safety tools exist, but that they consistently work in real-world conditions.
JBizNews Technology Desk
© JBizNews.com All Rights Reserved. Reproduction or distribution without written permission is prohibited.
