TikTok tests opt-in deepfake detection to protect creators from AI imposters
TikTok is testing a tool that scans AI-generated content on its platform for unauthorized use of real creators' faces. The feature is currently in a limited beta with a select group of US creators, making TikTok the second major video platform after YouTube to offer this kind of protection. With deepfake fraud rising and US legislators pushing the NO FAKES Act forward, the timing isn't accidental.
How it works
Creators who want to opt in must verify their identity through Jumio, a third-party service. The process requires a government-issued ID and a live selfie. TikTok spokesperson Zachary Kizer confirmed to Digital Trends that the platform does not retain copies of ID documents, and that biometric facial data is used solely to match against unauthorized AI-generated content — not stored for any other purpose.
Once enrolled, the tool monitors published content and alerts creators when a likely match is found. From there, creators can report the offending video or the impersonating account directly through the platform.
The opt-in difference
The choice to make enrollment voluntary is a deliberate one. Meta launched its Muse Image feature in early July 2026 with automatic opt-in, pulling in public Instagram photos to feed its AI image generator. The backlash was swift — Meta pulled the feature within three days, admitting it had "missed the mark," per Deadline. TikTok's model puts the decision entirely in the creator's hands, which should help it avoid a similar outcome.
YouTube moved first on likeness detection, expanding its tool from individual creators to celebrities and talent agencies in March 2026. TikTok's entry into the space gives creators on the platform a comparable option — though the beta remains US-only with no confirmed rollout date for the UK or broader international markets.
What's still unclear
Verification runs through Jumio, a US-based company, which raises unresolved questions about data residency for creators outside the United States. For UK-based users, post-Brexit data rules add another layer of complexity. TikTok has not disclosed accuracy figures for the detection system or how false positives will be handled. A wider rollout will likely depend on how those issues get resolved — and whether the NO FAKES Act becomes law, which would push platforms toward mandatory protections anyway.
For now, the tool is a meaningful step, but one with a narrow audience and several open questions still on the table.
