Seedance 2 pulled as it unexpectedly reconstructs voices accurately from face photos.
An AI model can now generate your voice from just a single picture of your face.
Deep Dive
The Seedance 2 AI model has been abruptly pulled from public access after users discovered it could accurately reconstruct a person's voice using only a single photograph of their face. The unexpected capability, which was not an advertised feature, raises immediate and serious privacy concerns about voice cloning and biometric security. The developers have not yet commented on the mechanism or provided a timeline for the model's return.
Why It Matters
This demonstrates a terrifying new frontier in AI-powered impersonation and identity theft.