Scarlett Johansson has accused OpenAI of using a voice that bears a strong resemblance to hers for its ChatGPT 4.0 system, dubbed “Sky,” despite her declining an offer to provide her voice for the AI.
In a statement to PEOPLE, Johansson disclosed that OpenAI CEO Sam Altman approached her in September 2023 with the opportunity to lend her voice to the AI, believing it could help bridge the gap between tech companies and creatives. However, citing personal reasons, she turned down the offer.
Nine months later, Johansson was surprised to find that the voice used in the released ChatGPT 4.0 demo, known as “Sky,” sounded remarkably similar to her own, causing confusion among her acquaintances and the public.
Johansson alleges that Altman hinted at the intentional similarity through a tweet mentioning “her,” a reference to a film where she voiced an AI system. She further claims that Altman tried to convince her to reconsider just days before the demo’s launch.
Feeling that her likeness was exploited, Johansson has taken legal action, seeking transparency and clarity regarding how OpenAI developed the “Sky” voice. She stresses the importance of safeguarding individual rights and likeness, particularly in an era of deepfakes and AI-generated content.
Acknowledging the concerns raised, OpenAI has temporarily halted the use of the “Sky” voice and outlined the voice selection process in a blog post. However, they maintain that the voice was never intended to resemble Johansson’s. Altman has issued an apology for the communication lapse, clarifying that the voice actor was chosen before any interaction with Johansson occurred.
This incident prompts ethical discussions about the utilization of AI-generated voices and the protection of individual rights in an increasingly AI-centric world.
