BEIJING — A Chinese court has ruled that using a person's voice to train an artificial intelligence model without consent constitutes a violation of personality rights.
The Shanghai No. 1 Intermediate People's Court on the 29th upheld a lower court ruling ordering platform operator A to pay 50,000 yuan (about 10 million won) in damages to a voice actor surnamed Wang over an AI-synthesized voice, according to Chinese outlet The Paper and other media on the 30th. It marks the first ruling in Shanghai in a voice rights dispute involving AI voice cloning.
The case began last year when Wang found a voice closely resembling their own at an online promotional event held by Company A. Suspecting the voice had been synthesized by AI, Wang filed suit. Company A acknowledged the voice was generated by its AI but said it could not confirm what material had been used for training because the employee responsible had left the company.
A forensic examination found that 90% of the AI-cloned voice was either similar or highly similar to Wang's actual voice. The court ruled that "using a person's voice as training data without their consent and imitating their tone, intonation and pronunciation to create a voice by which a specific individual can be identified constitutes a violation of voice rights." The company's failure to prove the legality of the source of its training data also weighed against it, the court said.
Disputes over AI voice cloning are mounting overseas as well. Japanese voice actor Kenjiro Tsuda filed a takedown suit against TikTok, saying more than 180 videos imitating his voice with generative AI had been posted on the platform. The case has prompted Japan's Ministry of Justice to consider protecting celebrities' voices under publicity rights.
In the United States, a lawsuit is under way alleging that AI voice startup Lovo used voice actors' voices to train AI without permission. A federal court in New York ruled last year that part of the actors' publicity rights claims could proceed.
The use of AI-cloned voices in financial fraud is also raising calls for regulation. In Italy this year, an affiliate of a bank was tricked into wiring about 95 million euros (about 145.9 billion won) to fraudsters who used AI to mimic the voices of its chief executive and a senior legal officer.
