- Johansson claims OpenAI used her voice without consent
- OpenAI denies intentional voice mimicry; pauses “Sky” voice use
- Johansson considers legal action over unauthorized AI voice use
Despite the CEO’s apparent suggestion in a social media post that the tech company had intentionally copied the actress’s intonation, the company has denied doing so.
According to Scarlett Johansson, she was “shocked” and “angry” when OpenAI purportedly replicated her voice without her permission for a new ChatGPT system.
The actress personally criticized Sam Altman, the company’s CEO, in a statement for suggesting that she was the voice of ‘Sky’ by publishing the word ‘her’ on X. This reference was to a film in which she portrayed an AI that a human fell in love with.
“In September of last year, I received an offer from Sam Altman to provide the voice for the current ChatGPT 4.0 system,” the 39-year-old Oscar nominee disclosed.
“He expressed his belief that by advocating for the system, I could foster a connection between tech companies and creatives and assist consumers in adjusting to the seismic shift between humans and AI.” He expressed his belief that my voice would provide solace to others.
She continued to explain that she ultimately declined his offer due to personal reasons. Nine months later, she noticed a striking similarity between the “Sky” voice and her own.
Johansson expressed her astonishment, anger, and disbelief upon hearing the released demo. She was perplexed that Mr. Altman would pursue a voice that was so strikingly similar to her own that even her closest friends and news outlets were unable to discern the difference.
Mr. Altman even suggested that the resemblance was deliberate by tweeting a single word, “her,” which is a reference to the film in which I portrayed a messaging system named Samantha, who develops an intimate relationship with a human.
Mr. Altman contacted my agent two days prior to the ChatGPT 4.0 demo’s release, urging me to reconsider. Before we could establish a connection, the system was operational prior to our ability to develop a connection.
Their actions forced me to hire legal counsel. The counsel wrote two letters to Mr. Altman and OpenAI, detailing their actions and requesting a detailed account of the process that generated the ‘Sky’ voice. As a result, OpenAI regretfully consented to the removal of the “Sky” voice.
“In a time when we are all struggling with deepfakes and the protection of our likeness, work, and identities, I am of the opinion that these are questions that require absolute clarity.” Transparency and the passage of suitable legislation will serve as a means of ensuring the protection of individual rights.
“Don’t miss out! Grab your free shares of Webull UK today!”
After users observed that the ChatGPT voice resembled the actress, OpenAI stated on Monday that it would “pause” its use.
OpenAI stated, “We have received inquiries regarding the selection of the voices in ChatGPT, particularly Sky.”
“We are working to pause Sky’s use while we address them.”
Through its ChatGPT service, the artificial intelligence (AI) corporation provides five voices that can articulate generated responses.
OpenAI previously denied that it had intentionally imitated Johansson’s voice and stated that it believed “AI voices should not deliberately mimic a celebrity’s distinctive voice.”
At the end of the previous year, the New York Times filed a lawsuit against OpenAI and its largest investor, Microsoft, for allegedly unlawfully using the newspaper’s articles to train and develop ChatGPT.
The lawsuit asserts that the AI text model is now competing with the newspaper as a reliable information source, which threatens the organization’s ability to provide this service.
