TikTok Tests AI Face Protection Tool to Help Creators Detect Deepfake Misuse
Identity Verification Is Required to Join the Beta
TikTok has tested a new feature that allows creators to detect unauthorized use of their face in AI-generated content. The video-sharing giant joined the likes of YouTube in rolling out the much-needed tool. However, the feature is only available to a select number of users in the United States.
Verification Is Key
The creators must verify their identity using Jumio, a service that requires a government-issued ID plus a live selfie, to activate the tool. When the scan finds possible matches, a notification pops up on the creator’s TikTok account, allowing them to report the offending content directly.
This approach prioritizes user’s consent, unlike what happened with Meta’s recently launched AI content tools. TikTok’s verification process means only those who have opted in – and provided facial recognition – would be able to detect potential deepfakes.
The Road Ahead Is Rocky
Besides the obvious concern about data privacy, several other issues need addressing before this technology can roll out worldwide. These include the storage and usage conditions of creators’ data not based in the U.S., how TikTok will handle false positives, and when the feature will be available in other countries. In addition, the timing could not be worse for TikTok. Lawmakers are pushing for the NO FAKES Act to restrict unauthorized use of people’s digital replicas. This act would force tech companies that use artificial intelligence to identify and remove unauthorized deepfake content and voice modulation.
From Labeling to Detection
The latest move illustrates how social media companies are shifting from labeling AI-generated content to focusing on tools that allow ordinary users to protect themselves from being impersonated in AI-generated content.
Get the latest tech news on WhatsApp
Comments are closed.