The acquisition happened earlier this year and covers the company’s small staff and intellectual property, The New York Times (NYT) reported Friday (May 15), citing sources familiar with the matter.
The report noted OpenAI’s revelation two years ago that it had developed the ability to replicate human voices with AI. The company said this tech was so advanced that it chose not to release it as a precaution. The Weights.gg acquisition, NYT added, shows the company has continued its work on voice AI, even as its position from two years ago has not changed.
Before the acquisition, the report added, Weights.gg provided a consumer app called Replay, that let users clone voices, including celebrities such as Taylor Swift and Samuel L. Jackson, both of whom have opposed the use of AI to clone their voices. (Swift has even gone as far as applying to the U.S Patent and Trademark Office to trademark her voice and likeness.)
The report also pointed out OpenAI’s history of copyright issues, including the ones that arose from the company’s now-shuttered video generation tool Sora.
While it’s not clear what OpenAI wants to do with the Weights team and technology, employees have been reassigned to work across different parts of the company, sources told NYT, adding that OpenAI is unlikely to put out a product similar to Weights.gg.
Advertisement: Scroll to Continue
In the meantime, the report continued, OpenAI has focused on weaving voice tech into other parts of the company. It recently began offering third-party developers information on how to use OpenAI’s application programming interface (API) to incorporate OpenAI’s voice technology into their apps and services.
These uses include things like providing real-time voice translation services or using voice commands to interact with AI agents, NYT said.
PYMNTS reported last year on research showing that AI has reached a point where cloned voices and genuine ones are indistinguishable.
“AI-generated voices are all around us now,” said lead author Nadine Lavan, senior lecturer at Queen Mary University of London, per a LiveScience report. “We’ve all spoken to Alexa or Siri, or had our calls taken by automated customer service systems. Those things don’t quite sound like real human voices, but it was only a matter of time until AI technology began to produce naturalistic, human-sounding speech.”
