ByteDance Suspends Seedance 2 Feature That Turns Facial Photos Into Personal Voices Over Potential Risks

Source: Slashdot

Seedance 2.0 Overview

China’s Bytedance has released Seedance 2.0, an AI video generator that can handle up to four types of input simultaneously: images, videos, audio, and text. Users may combine up to nine images, three videos, and three audio files—a total of twelve files. Generated videos run between 4 and 15 seconds (some reports cite up to 60 seconds) and automatically include sound effects or music.

Source: The Decoder

Suspension of the Facial‑to‑Voice Feature

The model’s performance proved so accurate that Bytedance was forced to block the facial‑to‑voice feature. The feature could generate highly accurate personal voice characteristics from facial photos alone, even without user authorization.

Details: TechNode article (Feb 10 2026)

Test Findings

In a recent test, Pan Tianhong, founder of the tech media outlet MediaStorm, discovered that uploading a personal facial photo caused the model to produce audio that was nearly identical to his real voice—without using any voice samples or authorized data.

Original report: TechNode article (Feb 10 2026)