A new study from researchers at New York University and Northeastern University claims that at least half of the safety features designed to protect children on major social media platforms don't work as promised. The research tested 86 protective features across Instagram, Snapchat, TikTok, and YouTube, finding failure rates of 50 percent or higher on each platform.

Researchers created dummy accounts to simulate children of various ages and adult users, then tested three scenarios: natural child usage, teens attempting to bypass protections, and malicious adults trying to contact minors. Features were marked as failures if they were too difficult to find, didn't function as advertised, or were missing entirely. The testing revealed that Snapchat allowed adult accounts to search for and message child accounts without restrictions, while Instagram permitted teen accounts to message adults they didn't follow without warnings.

TikTok was found suggesting anorexia-related searches to teen accounts. Spokespeople from Snap, Meta, and YouTube disputed the findings, though the New York Times reported it could replicate the results. The study adds pressure to social media companies already facing lawsuits from school districts and individuals claiming platform-induced harm.

Many countries are pursuing social media bans for children, with Australia recently doubling its maximum penalty for non-compliance. This research strengthens calls for stricter regulation as tech companies struggle to deliver effective age-appropriate safeguards.