OpenAI, the company behind ChatGPT, is venturing into the voice assistant industry and showcasing cutting-edge technology capable of replicating someone's voice.
However, the company decided not to make it available to the public at this time, citing safety concerns.
OpenAI Unveils Voice Engine Technology
The new Voice Engine technology was recently unveiled by the artificial intelligence company, shortly after filing a trademark application for the name.
According to the company, they have the ability to replicate someone's voice using only a 15-second audio recording of that individual speaking.
OpenAI has announced its intention to offer a preview of this technology to a select group of early testers, but it has decided against a widespread release for now due to concerns about potential misuse.
Recently, OpenAI introduced a cutting-edge tool that empowers individuals to effortlessly generate digital images by simply describing their desired visualizations. Then, it developed advanced technology capable of producing high-quality, full-motion video reminiscent of a Hollywood film.
Now, by uploading a recording of yourself along with a paragraph of text, the system can generate a synthetic voice that closely resembles your own.
The text can be in any language you prefer. For English speakers, it has the ability to replicate your voice in various languages such as Spanish, French, Chinese, and more.
AI Voice Generator Poses Threat to Safety Concerns
OpenAI is refraining from sharing the technology more extensively as it continues to delve into the potential risks it may pose.
A voice generator has the potential to contribute to the spread of disinformation on social media, similar to image and video generators. There is also the concern that it could enable individuals with malicious intent to assume false identities on the internet or during phone conversations.
The company expressed concern about the potential misuse of this technology, specifically in regards to compromising voice authenticators used for online banking accounts and other personal applications.
During an interview, Jeff Harris, a product manager at OpenAI, emphasized the importance of handling this sensitive matter with utmost care.
The company is currently investigating methods to implement watermarking on synthetic voices or incorporating restrictions to prevent the misuse of the technology with the voices of politicians or other well-known individuals.
In a recent development, OpenAI introduced Sora, a video generator, following a similar approach. The technology was demonstrated, but its release was not made public.
OpenAI is one of the companies that has successfully developed a new type of AI technology capable of generating synthetic voices with speed and ease. They feature tech giants such as Google, alongside start-ups like ElevenLabs based in New York.
The New York Times has filed a lawsuit against OpenAI and its partner, Microsoft, alleging copyright infringement related to the use of artificial intelligence systems that generate text.
These technologies can be utilized by businesses to create audiobooks, provide voice for online chatbots, or even develop an automated radio station DJ.
Over the past year, OpenAI has harnessed its technology to enable ChatGPT to communicate verbally. And it has always provided businesses with a wide range of voices that can be used for similar purposes. All of them were constructed using clips supplied by voice actors.
Authorities in New Hampshire are currently investigating a series of robocalls that were sent to a large number of voters right before the presidential primary.
These calls were particularly notable because they featured an AI-generated voice that sounded remarkably similar to President Joe Biden.
Several startup companies currently offer voice-cloning technology, with certain options available to the general public or exclusive to specific business customers like entertainment studios, AP News reported.
Join the Conversation