ByteDance Temporarily Disables Real-Person Reference Feature in Seedance 2.0
1 day ago / Read about 0 minute
Author:小编   

Recently, ByteDance's launch of Seedance 2.0, a cutting-edge video generation large model, has drawn considerable public interest. Tim, the founder of the self-media platform Film Storm, found during his assessment that by merely uploading a facial photo—without providing any accompanying audio files or specific instructions—the AI could autonomously generate a voice that closely resembled his own. This occurred without his prior authorization or any form of compensation. Such an incident has sparked widespread concerns regarding copyright infringement and privacy violations among the public.

In response to the escalating controversy, ByteDance swiftly suspended the real-person material reference capability within Seedance 2.0. The company stated that it is actively conducting urgent optimizations based on user feedback. Presently, the model does not support the use of real-person materials as the primary reference, and ByteDance has promised to reintroduce a more refined version of the feature in the future.