Hours after OpenAI announced that it would pause a ChatGPT voice that sounded like Hollywood film star Scarlett Johansson, the actress issued a statement saying she had a lawyer contact OpenAI CEO Sam Altman before the company took down its “Sky” voice option.
Johansson said in a statement that was reposted on X, formerly Twitter, that Altman contacted her last September and asked her to be the voice actress for its ChatGPT 4.0 system. She famously starred in a voice-only role in the movie Her, which is a romantic drama about a character who falls in love with an AI chatbot named Samantha. Altman told Johansson that she could “bridge the gap between tech companies and creatives and help consumers to feel comfortable with the seismic shift concerning humans and AI,” she said, adding that Altman told her that hearing her voice could be a comfort to people.
Johansson declined Altman’s offer, only to hear from friends, family and the public nine months later that the “Sky” voice sounded just like her.
“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,” she said, adding that Altman insinuated, with the one-word tweet “her,” that it was a conscious choice.
After her lawyers contacted OpenAI and Altman in two letters seeking exact detail on how they arrived at “Sky,” OpenAI “reluctantly agreed” to take down the voice, she said.
“In a time when we are all grappling with deepfakes and the protection of our own likeness, our own work, our own identities, I believe these are questions that deserve absolute clarity,” Johansson concluded.
Johansson’s description strikes right at the heart about debates over the ambiguity and potential misuses of of AI, chatbots, and voice and video cloning. She withheld consent from the company when it wanted to use her voice, and OpenAI instead ran with what is essentially her vocal clone. Her statement comes just days after Jan Leike, OpenAI’s head of alignment whose team focused on AI safety, resigned from the company.
In a statement on X, Leike said Friday that “safety culture and processes have taken a backseat to shiny products” at OpenAI and that he had been “disagreeing with OpenAI leadership about the company’s core priorities for quite some time, until we reached a breaking point.” His departure followed Ilya Sutskever, co-lead on OpenAI’s “Superalignment” team and chief scientist.
Earlier today, OpenAI posted on X that it had “heard questions” about how it chose the audio options for the chatbot and wanted to address them. The company also denied that the voice was based on Johansson’s.
“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,” the company wrote. It said it could not share the name of its voice actors for privacy reasons. Internet reactions were varied, but many poked fun at Sky’s tone.
In a statement to Fortune, Altman said the voice of Sky is not Johansson's "and it was never intended to resemble hers. We cast the voice actor behind Sky’s voice before any outreach to Ms. Johansson. Out of respect for Ms. Johansson, we have paused using Sky’s voice in our products. We are sorry to Ms. Johansson that we didn’t communicate better."
As Bloomberg reported, The Daily Show senior correspondent Desi Lydic said of GPT-4o in a segment last week, “You can really tell that a man built this tech.”
