Tests show Instagram continues to recommend sexual videos to accounts with ages listed as 13; Snapchat and TikTok did not produce the same sexualized content
Findings align with undisclosed internal research at Meta, which has pledged an ‘age-appropriate’ experience for teens …
Context & Ripple Effects
This is a continuation of a documented Instagram safety problem: a 2023 investigation found the service helped users find underage-sex-content networks, followed by tests showing Meta recommendations promoted underage sexual content despite the company’s child-safety efforts.
The findings also distinguish age assurance from feed safety. Instagram had already tested age-verification approaches for younger users, but these tests indicate that an account’s listed age did not reliably change what its recommendation system distributed.
First-order effects
- Meta faces fresh evidence that Instagram’s teen experience falls short of its age-appropriate pledge, specifically in recommendations delivered to accounts identified as 13.
- Snapchat and TikTok gain a comparative safety advantage in these tests because they did not return the same category of sexualized recommendations.
Second-order effects
- The contrast gives parents, advertisers, and safety advocates a concrete cross-platform benchmark, increasing pressure on Meta to demonstrate that ranking safeguards work in practice rather than merely exist as policy.
- It raises the importance of protections that operate at recommendation time, not only account setup—a gap made more visible after Instagram’s earlier testing of age-verification tools.
Third-order effects
- If repeated testing continues to find this gap, teen-safety evaluation will increasingly center on what platforms actively distribute, not just the content they remove or the age checks they offer.
- The broader structural shift is toward treating recommender systems as a distinct child-safety control surface, where platform claims can be compared through observable user outcomes.
The trend: Teen safety is becoming a test of recommendation-system behavior, with platforms increasingly judged on the feeds they generate for young users.