A 15-second recording is all that's needed to clone real voices. However, the threats to society mean OpenAI's new voice-cloning software will not be available for public use for the time being. — Photo: Hannes P. Albert/dpa
SAN JOSE: ChatGPT developer OpenAI is working on software to clone human voices, an innovation that could transform how films, shows and other content is consumed in foreign languages, but one that is also being seen as a threat to democracy.
"Voice Engine [...] uses text input and a single 15-second audio sample to generate natural-sounding speech that closely resembles the original speaker," the US company announced.
