Hollywood star Scarlett Johansson has said she was left "shocked" and "angered" after OpenAI launched a chatbot with an "eerily similar" voice to her own.
The actress said she had previously turned down an approach by the company to voice its new chatbot, which reads text aloud to users.
When the new model debuted last week commentators were quick to draw comparisons between the chatbot's "Sky" voice and Johansson's in the 2013 film Her.
OpenAI said on Monday that it would remove the voice, but insisted that it was not meant to be an "imitation" of the star.
However, Johansson accused the company, and its founder Sam Altman, of deliberately copying her voice, in a statement seen by the BBC on Monday evening.
“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," she wrote.
"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."
Set in the near future, 2013 film Her sees Joaquin Phoenix fall in love with his device's operating system, which is voiced by Ms Johansson.
The actress, who has been nominated for two Academy Awards, said she had been initially approached by Mr Altman about voicing the new chatbot in September.
"[Mr Altman] 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," Johansson wrote.
"He said he felt that my voice would be comforting to people."
But she eventually rejected the offer for personal reasons, she said.
Two days before the Sky chatbot was released, she added, Mr Altman contacted her agent, urging Johansson to reconsider her initial refusal to co-operate with the company.
Adding that she had been forced to hire lawyers, the actress said she had sent two legal letters to the company, to establish how the voice had been made.
"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," she wrote.
In a statement shared with the BBC by OpenAI, Mr Altman denied that the company had sought to imitate Johansson's voice.
"The voice of Sky is not Scarlett Johansson's, and it was never intended to resemble hers," he wrote.
"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.”
Separately, the firm said it was "working to pause" the voice while it addressed questions about how it was chosen in a post on X, formerly Twitter.
In its blog post, OpenAI said the five voices used by its chatbot had been sampled from voice actors it partnered with.
It comes just six months after actors agreed to end strikes which froze the entertainment industry amid calls for better pay and safeguards on the use of AI.
Ms Johansson took part in last year's industrial action, which was partly about how studios would use AI to imitate actors' faces and voices.
The deal struck with studios included reassurances that these would not be used without consent from actors.
"To use someone’s voice without permission feels particularly invasive at a time when distrust of AI and concern over its potential harms are rampant," said Dan Stein, head of AI voice licensing company Voice-Swap.
"Whether OpenAI trained their new Sky voice using audio from Scarlett Johansson or a sound-a-like, the fact remains that she refused permission and her identity was exploited regardless.
"It sets a dangerous precedent for copyright and consent if the most prominent company in the field behaves in this way."
OpenAI has been battling various legal challenges to how it uses copyrighted information available online.
In December, the New York Times said it planned to launch a lawsuit against the corporation over allegations that it had used "millions" of articles published by the media organisation to train its ChatGPT AI model.
And in September, authors George RR Martin and John Grisham also announced a plan to pursue a claim, over allegations their copyright had been infringed to train the system.
