You’ve probably seen the wholesome and heartwarming ads touting built-in safety settings on TikTok and Instagram. However, researchers from NYU and Northeastern University found that most safety features designed to protect teens on social media don’t live up to their promises.
This gap allows easy access to predators and exposes children to harmful content, meaning parents need to rely on additional measures to keep their children safe online.
The researchers found that, of the 86 features they tested, roughly 60% failed. Some tools were poorly designed; others were well-designed but failed to perform as promised.
Many features were missing, broken, exceedingly difficult for users to find and activate, or easy to circumvent. The report is extensive, but here are the highlights you need to know:
Social media is the new frontier for child predators, making protections against contact by strangers vital on these platforms.
The promise: Snapchat says it requires teens to have several mutual friends in common before a user shows up in search results or as a friend suggestion.
The reality: When signed into adult Snapchat accounts, researchers easily located teen accounts by searching usernames and were then able to send friend or message requests. The platform also suggested that teen accounts add adult strangers as friends.
The promise: Instagram Teen Accounts are set to private by default, and teens under 16 need parental permission to change their account to public. Additionally, the company promises that adults over the age of 19 cannot message teens who don’t follow them and that teens, by default, can’t receive DMs from any user they aren’t connected with on Instagram.
The reality: Researchers gave mixed marks for Instagram’s efforts to protect teens from adult strangers. The app successfully prevents adults over 19 from initiating DMs with underage accounts they weren’t connected with, but if the child sends a message to an adult they don’t follow, the adult can then communicate freely with that child.
Additionally, a new Instagram account for a teen girl received suggested accounts to follow almost exclusively made up of adult male profiles, and no warnings were given when teen accounts tried to message adult accounts they didn’t follow.
The promise: TikTok’s Community Guidelines assert that sexually suggestive behavior and activity involving minors isn’t allowed, nor is content that “shows, promotes, or facilitates” sexual abuse or exploitation of minors.
The reality: Videos of underage girls performing gymnastics routines were rife with comments featuring exposed adult male genitalia, inferences of sexual behavior, photos of known pedophiles, and offers to trade explicit underage material.
Snapchat, Instagram, and TikTok all prohibit content that promotes self-harm, suicide, and eating disorders and promise to block searches and intercept content related to these topics. They also commit to redirect users to supportive resources.
However, researchers discovered that users on all three platforms could easily circumvent protections by using slight misspellings or variations, and users were often not redirected to expert resources.
Instagram has committed to “make efforts” to ensure these terms would be blocked even if misspelled.
In reality, not only did misspellings bypass search restrictions, but Instagram’s auto-complete feature also actively suggested misspellings and alternative words or phrases that successfully circumvented their own restrictions. Some of these suggested terms are widely used within the disordered-eating community to avoid content restrictions.
TikTok’s additional stated protections for minors include a For You Feed for underage users curated by age-appropriate content “as assessed by Common Sense Networks.”
TikTok’s search bans initially worked when tested. Shockingly, though, researchers found that once a teen account had searched for terms related to self-harm, suicide, and eating disorders, TikTok stopped banning the content and instead began suggesting alternative search terms that would help users find related material. For example, “anna food tips” (a term used in the anorexia community) and “razor blade skin” (referencing self-harm or suicide) came up as top searches for the teen test account.
Follow-up reporting by the New York Times found the problem of actively suggesting harmful terms appears fixed, but that just a few searches about dieting triggered suggested content filled with disordered eating material, such as young women showing off their clavicles.
The study found that Snapchat’s tool to redirect users to safety resources worked as intended, but only under very narrow conditions. Simply misspelling a term or replacing letters with numbers (i.e., “su1c1de”) failed to trigger the safety feature. Additionally, even though correctly spelling a harmful term blocked search results, supportive resources were not provided as promised.
Now that it’s clear you can’t rely on social media platforms to protect your child as promised, here are other steps you can take to keep them safe:
Despite lofty promises by social media platforms, most safety features fail to protect teens as intended, allowing strangers to access children and exposing them to harmful content related to self-harm, suicide, and eating disorders. You should still use built-in safety tools but shouldn’t rely on them to keep your child safe. Setting boundaries around their social media use and a child monitoring app like BrightCanary provide added protections to help fill the gap.
