OpenAI Is Testing New AI Tool That Can Clone Human Voices With 15 Seconds of Audio Samples

Image Credit: Reuters

New Delhi: ChatGPT-maker OpenAI has been working on a new tool called Voice Engine. The AI-powered Voice Engine can analyze and clone a voice sample in just 15 seconds.

The latest AI tool can accurately produce a voice after listening to it for such a short duration. However, OpenAI is cautious about its widespread release due to concerns of potential misuse of synthetic voices.

The primary goal of this AI tool is to assist individuals with reading difficulties, including children, by generating lifelike voices for reading text aloud. Companies like Age of Learning utilise this technology to develop educational materials.

We're sharing our learnings from a small-scale preview of Voice Engine, a model which uses text input and a single 15-second audio sample to generate natural-sounding speech that closely resembles the original speaker. https://t.co/yLsfGaVtrZ

— OpenAI (@OpenAI) March 29, 2024

This tool can translate videos and podcasts into various languages while keeping the original speaker's voice, making it valuable for engaging with global audiences.

Despite its benefits, there are concerns about misuse, especially during election seasons. OpenAI says it wants to do more research into how it can protect tools like this from being used to spread misinformation and copy voices without consent.  

In a blog post by the company, it states “We recognize that generating speech that resembles people's voices has serious risks, which are especially top of mind in an election year". Moreover, “We are engaging with US and international partners from across government, media, entertainment, education, civil society and beyond to ensure we are incorporating their feedback as we build.”

Spotify has leveraged this technology in its pilot program to translate podcasts by renowned hosts like Lex Fridman. OpenAI highlighted additional uses, including expanding voice options for children's educational content.

OpenAI initiated Voice Engine development in late 2022. The technology has already powered preset voices for text-to-speech APIs and ChatGPT’s Read Aloud feature.