Actress Scarlett Johansson received an offer in September from OpenAI to use her voice in the ChatGPT-4o remake. Sam Altman, executive president of the company, told her that with her voice “she was going to close the gap between technology companies and creators and comfort users.” Johansson declined the invitation “after much thought,” but OpenAI went ahead with its plan to feature a voice similar to the actress: “My friends, family, and the general public noticed how the new system called Sky sounded like me,” says Johansson in a statement to NPR.
OpenAI announced this Monday, before the actress’s public message, the withdrawal of the Sky voice with a statement explicitly quoting Johansson: “We believe that AI voices should not deliberately imitate the particular voice of a celebrity. “Sky’s voice is not an imitation of Scarlett Johansson, but rather she belongs to a different professional actress who uses her own natural voice,” OpenAI said. Scarlett Johansson was the voice of the virtual assistant in the 2013 film Her, where the protagonist falls in love with her. Altman himself tweeted “her” during the presentation of the ChatGPT-4o version.
“When I heard the new voice, I was shocked, I was angry and I couldn’t believe it,” says the actress, who claims that Altman’s X message with the film’s title is proof that Altman’s intentions were to copy his voice without your permission. Johansson also reveals that two days before ChatGPT-4o’s presentation last Monday, Altman contacted his agent: “Before we could respond, the system was already out there,” says the actress.
The controversy reveals two problems with artificial intelligence and Altman’s attitude as head of OpenAI. Since the emergence of AI, creators have seen how these systems have fed off decades of human artistic work to produce artificial creations without their permission. Johansson’s case is so far the most egregious: Altman even asked her for collaboration, but he didn’t care when the actress said no. “At a time when we are all dealing with deepfakes and the protection of our image, our work, our identities, I believe that these issues deserve complete clarity,” says Johansson, who adds: “I hope that a resolution in the form of transparency and the adoption of appropriate legislation will help ensure that the rights individuals are protected.”
The second problem is Altman’s attitude since his departure and triumphant return in November as CEO of OpenAI. Last week several of his main collaborators left the company due to disagreements. One of them, Jan Leike, in charge of analyzing risks in OpenAI’s work, regretted that the company focuses on flashy products.
The conversation with Johansson reveals how Altman despises the obstacles that limit his goals. Especially when he had promised not to humanize his chatbots nor encourage emotional relationships with them. The effort to achieve the voice of Her indicates just the opposite: few scruples to advance on his path. The outgoing board said it had fired Altman because he had not been “consistently frank” in his dealings with them. Now Johansson has given the first public test.
You can follow The USA Print in Facebook and x or sign up here to receive our weekly newsletter.
