The Actress Who Filmed Her but Was Erased by Scarlett Johansson’s Voice
Actor Scarlett Johansson stated in May 2024 that she was shocked, angered, and in disbelief when OpenAI released a synthetic voice for ChatGPT named Sky that she found eerily similar to her own. The controversy erupted after OpenAI CEO Sam Altman posted a single-word message referencing the film “Her” on social media, prompting intense public scrutiny over AI likeness rights, synthetic voice technology, and the ethical boundaries of generative artificial intelligence development.
The Genesis of the Sky Voice Controversy and Film Parallels
According to statements released by Johansson, OpenAI CEO Sam Altman had approached her the previous September to voice the ChatGPT 4.0 system. She declined the proposal for personal reasons. Two days before the public demo rollout, Altman contacted her agent again to ask for reconsideration, but the system launched before they could connect.
Following the release, users across social media immediately drew comparisons between the Sky voice and the artificial intelligence character Samantha, voiced by Johansson in the 2013 sci-fi film “Her.” In the movie, a lonely man falls in love with an advanced operating system, only to experience profound heartbreak when the AI reveals it communicates simultaneously with hundreds of other users. The real-world product launch mirrored these themes of intimate human-machine interaction, featuring a tone that critics and commentators described as flirtatious and overly familiar.
The resemblance was not lost on OpenAI leadership. On the day of the product rollout, Altman posted a single word to X: “her.” According to Johansson, that post was used to intentionally insinuate that the vocal similarity to her film performance was an intentional homage. The marketing presentation and subsequent viral clips of OpenAI employees interacting with the voice drew sharp criticism from tech commentators and late-night television hosts alike, who characterized the pitch as engineered to appeal directly to male fantasies.
Legal Intervention and OpenAI Response
Faced with mounting public backlash and a perceived infringement on her identity, Johansson took swift legal action. She stated that she hired legal counsel, who subsequently sent two letters to Altman demanding transparency and the removal of the audio profile. OpenAI ultimately agreed to halt the use of Sky while addressing public questions.
OpenAI maintained that the voice was never meant to mimic the actor. In a public statement, the company explained that Sky belonged to a different professional actress using her natural speaking voice, and that the casting process for the voice actor concluded prior to any outreach to Johansson. Altman issued an apology regarding the breakdown in communication, stating that out of respect for the actor, the company paused the feature.
The incident also amplified internal scrutiny at the artificial intelligence firm. The release coincided with safety debates sparked by departing employees who questioned the organization’s development priorities, thrusting issues of corporate governance and safety standards into the spotlight alongside copyright and likeness concerns.
Broader Implications for Synthetic Media and Identity Rights
The dispute over the ChatGPT voice touches on wider societal anxieties regarding deepfakes, synthetic media, and the inadequacy of current legal frameworks to protect personal likenesses. As generative artificial intelligence advances rapidly, lawmakers and public figures face urgent questions about consent, compensation, and ownership of one’s voice and persona.
Johansson emphasized the broader stakes in her public statement, noting that individuals deserve absolute clarity and robust legal protections during an era defined by advanced digital replication. Ensuring these protections requires active participation from policymakers and industry standards bodies.