Actress Scarlett Johannson was “shocked” and “angered” upon hearing ChatGPT’s “Sky” voice, which some had likened to her own. Johannson famously voiced a fictional digital assistant in the 2013 film Her.
In a statement released Monday, obtained by NBC News, Johannson felt the comparisons were apt and said the Sky voice was “eerily similar” to hers.
Crucially, she added that she’d received an offer last September from OpenAI CEO Sam Altman to voice Sky — and that she declined the offer.
“He told me that he felt that by my voicing the system, I could bridge the gap between tech companies and creatives and help consumers to feel comfortable with the seismic shift concerning humans and AI,” Johannson said. “He said he felt that my voice would be comforting to people.”
Prior to Johannson’s statement, Open AI made this claim in a post Sunday: “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.”
Nevertheless, OpenAI announced that same day that it would stop using Sky in its ChatGPT product until it resolved lingering issues about its origins.
“We’ve heard questions about how we chose the voices in ChatGPT, especially Sky,” the company wrote in a post on X Sunday. OpenAI is “working to pause the use of Sky” while it addresses those questions.
OpenAI drops ChatGPT’s ‘Sky’ voice after Scarlett Johansson ‘Her’ comparisons
“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 in her statement.
The Sky voice has been around since OpenAI first released Voice Mode in September of 2023 as one of five options to choose from in addition to “Breeze”, “Cove”, “Ember”, and “Juniper”. But OpenAI’s announcement of GPT-4o, which brings these voice capabilities directly into the model, exclusively uses Sky, thus casting a spotlight on that particular voice.
During the GPT-4o live demo last week, OpenAI CEO Sam Altman picked up on the comparisons to Scarlett Johansson. He posted the single word “her” on X.
Johansson said in her statement that Altman had contacted her agent, urging her to reconsider his offer, two days prior to the demo release. She said they never connected before the system debuted, and that she hired legal counsel as a result.
Johansson claimed that OpenAI “reluctantly” agreed to pause the Sky voice, following her counsel’s request for detailed information about how Sky was created.
“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 said. “I look forward to resolution in the form of transparency and the passage of appropriate legislation to help ensure that individual rights are protected.”
Responding to NBC News’ request for comment, Altman stated that Sky’s voice was not Johansson’s, and that “it was never intended to resemble hers.”
“We cast the voice actor behind Sky’s voice before any outreach to Ms. Johansson,” said Altman. “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.”
Cecily Mauran contributed to this story.
UPDATE: May. 21, 2024, 2:56 p.m. AEST This article has been updated with OpenAI CEO Sam Altman’s comment to NBC News.
