ChatGPT's Voice Revelation: Actress' Agent Claims No Imitation, OpenAI's Safety Board Disbands

New revelations have emerged regarding Scarlett Johansson's plagiarism accusation against ChatGPT's parent company, OpenAI. An agent for the actress who voiced "Sky" for the AI chatbot claims her client was not instructed to imitate Johansson's voice. Meanwhile, OpenAI's AI safety task force has been disbanded.

In a surprising turn of events, a new agent for the actress who voiced "Sky" for ChatGPT has denied that her client was tasked with imitating Scarlett Johansson's voice. According to The Washington Post, the agent stated that neither Johansson nor her movie "Her" were mentioned during her client's audition for the role.

The agent revealed that the name "Sky" was chosen to evoke a "cool, airy, and pleasant sound." Audio recordings of the actress's voice test reportedly sounded identical to the AI-generated Sky voice.

Johansson had previously accused OpenAI of copying her voice after turning down an offer to be a human voice for ChatGPT's Voice Mode. After being contacted in September by CEO Sam Altman, she rejected the offer "for personal reasons." However, she was later shocked to hear the "Sky" voice, which she believed sounded eerily similar to her performance in "Her."

Johansson has since hired legal counsel and has suggested that her lawyer's demands for an explanation led OpenAI to remove the "Sky" voice on Sunday.

Altman has denied that "Sky" was meant to resemble Johansson's voice, stating that the voice actor was cast before any contact was made with her. He expressed regret for not communicating better with Johansson and announced that OpenAI had paused using the Sky voice out of respect.

Joanne Jang, a product lead at OpenAI, revealed that Altman was not heavily involved in the casting process for ChatGPT's Voice Mode due to a world tour.

The unidentified voice actress expressed her discomfort with the backlash, as it was her natural voice being compared to Johansson's. She acknowledged the inevitability of voice-over actors embracing the future of AI technology.

Representatives for Johansson have not yet responded to requests for comment.

In a separate development, OpenAI has disbanded its AI safety task force. The company stated that the board had "completed most of its initial goals" and that its responsibilities would be integrated into other teams.

This move has raised concerns about the prioritization of AI safety within OpenAI. However, the company has emphasized that AI safety remains a core value and that ongoing work in this area will be folded into other teams' mandates. Representatives for Johansson and OpenAI did not immediately respond to requests for comment.