Artificial intelligence company OpenAI, the creators of ChatGPT, revealed its work on a new AI tech system, Voice Engine, that can recreate a specific person’s voice using just a 15-second audio recording. The produced voice can also be spoken in languages that aren’t a person’s native tongue.
OpenAI has not yet shared the technology behind the system because the company is trying to better understand the potential dangers it creates. Not only could it be part of spreading disinformation online, but it might also allow criminals to pose as others online or on the phone. Most worrisome, OpenAI says, is the potential that it could be used to bypass voice authenticators that allow access to bank accounts and other private applications.
The company said it was particularly worried that this kind of technology could be used to break voice authenticators that control access to online banking accounts and other personal applications.
“This is a sensitive thing, and it is important to get it right,” said OpenAI product manager Jeff Harris. The company is looking into how to “watermark” synthetic voices or adding controls that could prevent peoples from impersonating the voice of public figures, including politicians.