ChatGPT: Scarlett Johansson left ‘furious’ over chatbot imitation

Hollywood star Scarlett Johansson said she was “shocked” and “outraged” after OpenAI launched a chatbot with a voice “eerily similar” to hers.

The actress said she had previously rejected the company’s approach to voicing its new chatbot, which reads text aloud to users.

When the new model, called Sky, debuted last week, commentators were quick to draw comparisons between the chatbot’s tone and Johansson’s in the 2013 film Her.

OpenAI said on Monday it would remove the audio, but insisted it was not intended as an “imitation” of the star.

However, Johansson accused the company, and its founder, Sam Altman, of deliberately imitating her voice, in a statement seen by the BBC on Monday evening.

“When I heard the released demo, I was shocked, angry, and in disbelief that Mr. Altman would follow a voice that sounded eerily similar to mine,” she wrote.

“Mr. Altman even hinted that the resemblance was intentional, tweeting one word ‘she’ — a reference to the movie in which I voiced a chat system, Samantha, who forms an intimate relationship with a human.”

The 2013 film Her is set in the near future, where Joaquin Phoenix falls in love with his device’s operating system, voiced by Ms. Johansson.

The actress, who has been nominated for two Oscars, said Mr. Altman initially approached her about voicing the new chatbot in September.

“[Mr Altman] “He told me that he felt that by voicing the system, I could bridge the gap between technology companies and creators and help consumers feel comfortable with the seismic shift in relation to humans and AI,” Johansson wrote.

“He said he felt like my voice would be comforting to people.”

But she said she ultimately rejected the offer for personal reasons.

She added that two days before the Sky chatbot was launched, Altman called her agent and urged Johansson to reconsider her initial refusal to cooperate with the company.

The actress added that she had to hire lawyers, and said that she sent two legal letters to the company, explaining how to produce the sound.

“At a time when we are all grappling with deepfakes and protecting our image, our work, and our private identities, I believe these are questions that deserve absolute clarity,” she wrote.

In a statement OpenAI shared with the BBC, Altman denied that the company had sought to imitate Johansson’s voice.

“Sky’s voice is not Scarlett Johansson’s voice, and was never intended to sound like hers,” he wrote.

“We cast a voice actor behind Skye’s voice prior to any communication with Ms. Johansson. Out of respect for Ms. Johansson, we have temporarily stopped using Skye’s voice in our products. We apologize to Ms. Johansson that we did not communicate better.”

Separately, the company said it was “pausing” the sound while responding to questions about how it was selected in a post on X, formerly Twitter.

In its blog post, OpenAI said that the five voices used by its chatbot were sampled from the voice actors it partnered with.

It comes just six months after actors agreed to end strikes that froze the entertainment industry amid calls for better wages and safeguards over the use of artificial intelligence.

Ms. Johansson participated in the industry event last year, which was partly about how studios are using artificial intelligence to mimic actors’ faces and voices.

“Using someone’s voice without permission is especially violative at a time when mistrust of AI and concern about its potential harms are rampant,” said Dan Stein, president of AI voice licensing company Voice-Swap.

“Whether OpenAI trained Sky’s new voice using a voice from Scarlett Johansson or a similar voice, 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 industry acts in this way.”

OpenAI faces several legal challenges related to how it uses copyrighted information available online.

In December, The New York Times said it planned to file a lawsuit against the company over allegations that it used “millions” of articles published by the media organization to train its ChatGPT AI model.

In September, authors George R.R. Martin and John Grisham also announced a plan to pursue a lawsuit, over allegations that their copyright to the training system was infringed.