Last week, OpenAI revealed a new conversational interface for ChatGPT with a expressive and synthetic voice strikingly similar to that of the AI assistant played by Scarlett Johansson in the sci-fi film Her-only to suddenly turn off the new voice this weekend.
On Monday, Johansson released a statement claiming to have forced the reversal, after his lawyers asked OpenAI to clarify how the new voice was created.
Johansson statement, relayed to WIRED by her publicist, claims that OpenAI CEO Sam Altman asked her last September to provide the new voice for ChatGPT but that she refused. She describes being stunned to see the company introduce a new voice for ChatGPT last week that sounded like her anyway.
“When I heard the demo, I was shocked, angry and incredulous that Mr. Altman could pursue a voice that sounded so eerily like mine that my closest friends and the media couldn't tell the difference.” , we can read in the press release. He notes that Altman seemed to encourage the world to connect the demo to Johansson's performance by tweeting “her», in reference to the film, on May 13.
Johansson's statement said his agent was contacted by Altman two days before last week's demo to ask him to reconsider his decision not to work with OpenAI. After seeing the demo, she says she hired legal counsel to write to OpenAI to ask for details about how the new voice was created.
The statement claims this led to OpenAI's announcement on Sunday in a post on that she had decided to “suspend the use of Sky”, the name given by the company to the synthetic voice. The company also published a blog post describing the process used to create the voice. “Sky's voice is not an impersonation of Scarlett Johansson but belongs to another professional actress using her own natural voice,” the post said.
Sky is one of several synthetic voices OpenAI gave to ChatGPT last September, but at last week's event she displayed a much more realistic intonation with emotional cues. The demo saw a version of ChatGPT powered by a new AI model called GPT-4o appear flirting with an OpenAI engineer in a way that many viewers found reminiscent of Johansson's performance in Her.
“Sky's voice is not Scarlett Johansson's, and it was never intended to sound like her,” Sam Altman said in a statement provided by OpenAI. He claimed the voice actor behind Sky's voice was hired before the company contacted Johannsson. “Out of respect for Ms. Johansson, we have stopped using Sky's voice in our products. We apologize to Ms. Johansson for not having communicated better.
The conflict with Johansson adds to OpenAI's existing battles with artists, writers and other creatives. The company already defends a number of prosecutions alleging that it improperly used copyrighted content to train its algorithms, including lawsuits from the New York Times and authors including George RR Martin.
Generative AI has made it much easier to create realistic synthetic voices, creating new opportunities and threats. In January, New Hampshire voters were bombarded with robocalls featuring a fake voicemail from Joe Biden. In March, OpenAI announced that it had developed technology that could clone someone's voice from a 15-second clip, but the company said it would not release the technology because of how it could be misused.
Updated 5/20/2024, 9 p.m. EDT: This article has been updated with comment from OpenAI CEO Sam Altman.