OpenAI founder Sam Altman suddenly faces a more formidable foe than any members of his company’s board of directors: Hollywood star Scarlett Johansson.
The actor released a statement saying that Altman in September had asked her to voice ChatGPT’s text-to-speech product because he found her voice “comforting,” but that she declined the offer, according to the New York Times. Nine months later, when ChatGPT introduced its voice product, one option sounded so much like Johansson herself that she said she was “shocked, angered and in disbelief that Mr. Altman would pursue a voice that sounded so eerily similar to mine that my closest friends and news outlets could not tell the difference.”
Johansson added in her statement to the Times that she hired legal counsel to ask Altman how the company had created the the voice, which OpenAI calls Sky. Altman has stated publicly that the 2013 movie “Her,” starring Johansson as a voice robot, is his favorite film.
The company’s so-called voice mode function, launched in 2023, lets users interact with ChatGPT, widely used “generative” artificial intelligence software, by speaking to it. It also acts as a text-to-speech tool that reads answers to user prompts in one of five different voice options.
Although OpenAI said Sky is not Johansson’s, the company said it is pausing its use to respond to user questions about how it chooses and samples the voices in ChatGPT.
“We are working to pause the use of Sky while we address them,” the company said in a statement on X (formerly Twitter).
In a blog post late Sunday, OpenAI elaborated on how it comes up with its different voices, reiterating that the Sky option uses the voice of an actor who is not Johansson.
“We believe that AI voices should not deliberately mimic a celebrity’s distinctive voice — Sky’s voice is not an imitation of Scarlett Johansson but belongs to a different professional actress using her own natural speaking voice. To protect their privacy, we cannot share the names of our voice talents,” the company said in the post.
OpenAI worked with casting professionals and voice actors, ultimately whittling the five voices down from a pool of 400 options. The actors are paid as long as their voices are used in ChatGPT’s products, the company noted. Winners were selected based on a variety of criteria, including a sense of timelessness and how easy the voices are to listen to.
Johansson on Monday said via her publicist that Altman had reached out to her last September about voicing the system but said she “declined the offer.” She said Altman reached out to her agent again two days before the voice demo went live “asking me to reconsider,” but the voice was released “before we could connect.”
“When I heard the released demo, I was shocked, angered and in disbelief that Mr. Altman would pursue a voice that sounded so eerily similar to mine that my closest friends and news outlets could not tell the difference,” Johansson said. “Mr. Altman even insinuated that the similarity was intentional, tweeting a single word ‘her’ — a reference to the film in which I voiced a chat system, Samantha, who forms an intimate relationship with a human.”
Johansson said she had lawyers contact Altman and OpenAI “asking them to detail the exact process by which they created the ‘Sky’ voice.” It was after these demands that the voice was taken down, Johansson said.
Some users lamented the company’s decision to temporarily disable Sky, calling it “the best” or their “favorite” voice. It remains unclear why, if the voice was sampled from a paid actor, the company felt it necessary to take it down. OpenAI did not indicate when the voice will become available again.
Source: Scarlett Johansson was “shocked, angered” by OpenAI’s ChatGPT voice that sounds like her