Child safety tools on the biggest social media platforms often fall short. A report released Monday shows many advertised protections on Instagram, Snapchat, TikTok and YouTube simply do not deliver. Researchers from New York University and Northeastern University spent months testing dozens of these features. What they found was sobering.
Some safeguards were missing. Others proved broken or easy to dodge. Still more were buried so deep in menus that parents and teens never saw them. The New York Times replicated several of the experiments. The results held up.
Snapchat, for instance, continued to let adults send message requests directly to accounts held by children. The app also suggested that young users become friends with adult strangers. On Instagram, teen profiles received prompts to connect with unknown men. TikTok, despite public promises to crack down on content that promotes eating disorders, still surfaced searches such as “how to pretend to eat your food.”
These lapses arrive at a charged moment. Tech companies face lawsuits that could cost them billions. Several countries have already banned social media use for anyone under 16. Lawmakers on Capitol Hill have scheduled testimony from industry leaders next month. The gap between corporate assurances and actual outcomes has never looked wider.
Platforms insist they have grown safer over the years. They point to new parental controls and screen-time limits. Yet the NYU and Northeastern findings paint a different picture. “The language often implied that the platforms were doing much more to prevent harm than they actually were,” said Lexie Matsumoto, a graduate student at New York University who worked on the project. The New York Times quoted her directly.
The report, available at cybersafetyresearch.org/broken_buried_missing.pdf, catalogs case after case where promised protections crumbled under real-world use. Features designed to block harmful content failed to catch it. Age gates proved trivial to bypass. Default settings pushed young users toward risky interactions.
But the problem runs deeper than faulty code. Industry design choices reward engagement above all. Algorithms surface material that keeps eyes glued to screens, even when that material damages developing minds. Internal company documents unsealed in ongoing litigation reveal executives long understood these dynamics. They knew certain product features fueled compulsive use among minors. Yet those features stayed.
Recent data only sharpens the alarm. A June 2026 EU Eurobarometer survey found young people average 4.5 hours online on school days and more than six hours on weekends. Nearly one in three adolescents reported feeling stressed, sad or left out because of social media. One in four had encountered harmful content ranging from hate speech to violent material. Nine out of ten described at least one physical symptom tied to heavy screen time, from headaches to trouble concentrating. The European Commission published the survey on June 16.
Population-level effects now appear. The 2026 World Happiness Report devotes an entire chapter to evidence that social media harms adolescents on a scale large enough to shift mental health trends across entire age groups. Direct harms such as sextortion and cyberbullying combine with indirect ones like rising depression and anxiety. Researchers present seven lines of evidence that current use patterns fail any reasonable safety test. The full chapter is at worldhappiness.report.
Legislators have taken notice. Utah passed eight separate bills since 2023 targeting addiction, age verification, predatory algorithms and adult content. The state now ranks first nationally for online child protection according to the Childhood Index. Other states and nations follow. Australia began enforcing a social media minimum age of 16 in December 2025; early compliance data shows more than 310,000 additional accounts blocked by March 2026. The eSafety Commissioner’s update is at esafety.gov.au.
Canada, the United Kingdom and members of the European Union weigh similar restrictions. Many proposals include carve-outs: platforms can avoid outright bans if they demonstrate strong, independently verified safety systems. Yet the NYU report suggests most current systems would not qualify. And. That raises hard questions about enforcement and technical feasibility.
Parents stand caught in the middle. Many want better tools, not blanket prohibitions. A 2025 Center for Democracy and Technology study found both teens and caregivers desire safety features that respect privacy and avoid overreach. The gap between policy ideas and lived experience remains wide. The report sits at cdt.org.
UN human rights officials warn against shifting the entire burden onto families. In May they released guidelines titled “Getting Children’s Safety Online Right.” The document calls for safety by design, mandatory oversight and corporate accountability rather than parental workarounds. Volker Türk, the UN High Commissioner for Human Rights, stressed that states and companies must act together. Details appear at ohchr.org.
Tech executives will soon face Congress. Their prepared remarks will likely emphasize incremental progress and investment in new moderation technology. Yet the independent research keeps piling up. A September 2025 analysis cited in recent coverage found 64 percent of teen safety tools ineffective, defunct or easily bypassed. Meta disputed those numbers. The pattern, however, repeats across platforms.
Disappearing messages on Snapchat hide conversations from parents. TikTok’s recommendation engine pushes vulnerable users toward harmful communities. Instagram’s explore page can amplify body-image anxiety then feed related disorder content at triple the normal rate, according to an internal Meta study reviewed by Reuters in 2025.
So the evidence converges from multiple directions. Academic tests. Government surveys. Lawsuit discovery. International reports. Each shows the same mismatch: companies promote safety while their products continue to expose children to predictable harms.
Whether new laws will close that gap remains uncertain. Age verification technology brings privacy risks of its own. Enforcement across borders is difficult. And many young users already know how to circumvent restrictions. Yet the status quo clearly fails basic tests of responsibility.
One fact stands out. Children cannot consent to contracts, cannot buy alcohol, cannot drive. Society long ago decided certain activities require maturity. The question now is whether social media belongs on that list or whether platforms can be forced to redesign their products so that maturity is not required for safe use.
The NYU and Northeastern researchers have given policymakers fresh ammunition. Their work shows that many existing safety features amount to little more than marketing language. Real protection will demand more than promises. It will require changes to core architecture, transparent measurement and genuine accountability. Anything less leaves children exposed while the companies continue to profit.
WebProNews is an iEntry Publication