OpenAI Unveils Voice Cloning Technology
OpenAI has revealed its Voice Engine technology, capable of cloning a person’s voice with just 15 seconds of recording. Despite its potential, the company has opted not to release it publicly yet due to safety concerns. This decision comes amid worries about misuse, particularly in sensitive contexts like elections. OpenAI plans to offer previews to early testers but will not widely release the technology at this time.
The move follows investigations into AI-generated robocalls mimicking public figures. While some startups offer similar technology, OpenAI emphasizes ethical use, requiring testers to obtain consent and disclose AI-generated voices. The company’s trademark application suggests intentions to enter the speech recognition and digital voice assistant market, potentially competing with products like Amazon’s Alexa.