What Tiktok Actually Requires: Accepted Documents and Selfie Checks for Age Verification
Facial age estimation reduces document exposure, but it brings technical trade-offs. Independent audits of AI age estimation systems demonstrate that error rates fluctuate across demographics. Facial estimation tools maintain an average margin of error of roughly 1.2 to 1.5 years for adults aged 20 to 40. That margin widens significantly among teenagers passing through puberty, where hormonal developments alter bone structure unevenly.
Lighting conditions also distort the engine's readouts. Dim lighting, backlight glare, or thick cosmetic makeup can prompt the estimation system to register an adult as a minor.
When a user sits directly on an age threshold, such as an 18-year-old whose structural features read as 16, algorithmic verification frequently stalls. For creators whose livelihoods depend on unhindered account access, relying exclusively on video selfies can lead to repeated, false suspensions. In those edge cases, submitting a redacted physical ID document provides a faster path to restoring full access.