A video surfaced in the feed of millions of social media users in February 2025. Scarlett Johansson was seen sporting a white T-shirt with the name Kanye, a Star of David, and a hand giving the middle finger. She was accompanied by what looked to be Jerry Seinfeld, Adam Sandler, Mila Kunis, Jack Black, and perhaps fifteen other Jewish celebs, all of whom were wearing the same shirt and denouncing Kanye West’s antisemitic rants on X. The video’s message was clear. The famous people in it weren’t real. They had all refused to show up.
Digital marketers at Gitam BBDO, an Israeli creative agency, produced the film by simulating the faces of well-known Jewish public figures using generative AI. With the faces of people most closely linked to Jewish public life in America, it was intended as counter-speech, a reaction against hatred. It became well-known. “There is a 1000-foot wave coming regarding AI,” Scarlett Johansson told People magazine, “that several progressive countries, not including the United States, have responded to in a responsible manner.”
This statement changed the focus of the conversation from the Kanye West story to the one beneath it. The US government, according to her, is “paralyzed when it comes to passing legislation that protects all of its citizens against the imminent dangers of AI.” The lack of action, she said, was “terrifying.” Because it came with a particular context—Johansson was not new to this argument—the statement landed differently than a normal celebrity social media post. It has been her home for years.
Important Information
| Field | Details |
|---|---|
| The Deepfake Video (February 2025) | An AI-generated video featuring Scarlett Johansson’s likeness, alongside AI versions of approximately 20 other Jewish celebrities including Jerry Seinfeld, Adam Sandler, Mila Kunis, and David Schwimmer, went viral after being created by Israeli digital marketing professionals Ori Bejerano and Guy Bar of agency Gitam BBDO as a response to Kanye West’s antisemitic social media posts; Johansson publicly condemned both the antisemitism and the unauthorised use of her likeness |
| Johansson’s Statement | Told People magazine: “We must call out the misuse of AI, no matter its messaging, or we risk losing a hold on reality. There is a 1000-foot wave coming regarding AI that several progressive countries, not including the United States, have responded to in a responsible manner. I urge the US government to make the passing of legislation limiting AI use a top priority.” |
| The OpenAI Voice Dispute (May 2024) | OpenAI released an AI assistant voice called “Sky” that Johansson and her team described as sounding “eerily similar” to her voice; Johansson said she had declined OpenAI’s prior collaboration requests; OpenAI denied Sky was an imitation of Johansson but suspended the voice; SAG-AFTRA supported Johansson’s position; no formal lawsuit was publicly confirmed as filed |
| The NO FAKES Act | Bipartisan legislation (Nurture Originals, Foster Art, and Keep Entertainment Safe) reintroduced to both the US Senate and House on April 9, 2025 by Senators Chris Coons (D-DE), Marsha Blackburn (R-TN), Amy Klobuchar (D-MN), and Thom Tillis (R-NC); would establish a federal right for every American to their voice and visual likeness; require platforms to promptly remove unauthorised deepfakes; supported by SAG-AFTRA, RIAA, MPA, Sony Music, Warner Music Group, Universal Music Group, YouTube, and OpenAI |
| Prior Federal Attempts | Deepfake Report Act of 2019; Deepfake Task Force Act (2021); Preventing Deepfakes of Intimate Images Act (2023); Deepfakes Accountability Act (2023) — none were enacted into law; current US law offers only a patchwork of state-level publicity rights protections |
| Industry Support Coalition | The Walt Disney Company, IBM, YouTube, OpenAI, SAG-AFTRA, the Recording Academy, Universal Music Group, Sony Music, Warner Music Group, the Motion Picture Association and dozens of other organisations have publicly endorsed the NO FAKES Act |
Johansson and her legal team claimed that Sky, an AI voice assistant published by OpenAI in May 2024, uncannily resembled Johansson’s voice, which she had used to represent an AI companion in the 2013 movie Her. She said that she had previously declined OpenAI’s requests to contribute her real voice to the project. Despite denying that Johansson was the model for Sky, OpenAI halted the voice. Just prior to Sky’s release, the CEO of the firm cited a line from Her in a tweet, which Johansson claimed was hard to take as coincidental. SAG-AFTRA supported her stance. The controversy had already done a lot of work in Washington just by existing, even though no formal lawsuit was officially confirmed.
The February deepfake video and the OpenAI voice dispute combined to set the stage for something Congress had been debating for years but had never finished. The NO FAKES Act, formerly known as the Nurture Originals, Foster Art, and Keep Entertainment Safe Act, was reintroduced on April 9, 2025, by a bipartisan coalition of senators led by Chris Coons and Marsha Blackburn, along with Amy Klobuchar and Thom Tillis.
Every American would have a federal right to their voice and physical likeness under the measure. Platforms would have to quickly eliminate unauthorized deepfakes. It would establish a notice-and-takedown system based on current copyright legislation. The coalition’s support was what set this version of the measure apart from those that had stalled in earlier sessions. The names SAG-AFTRA, RIAA, and Motion Picture Association were not new.
This time, the list of endorsements also included OpenAI and YouTube. OpenAI’s Anna Makanju stated that the company was “pleased to support the NO FAKES Act.” Chris Lehane, CEO of OpenAI, described it as evidence of what happens when the creative and tech sectors “come together.” A corporation that was involved in one of the most well-known AI likeness debates in recent memory is now supporting the legislation that resulted from that disagreement, which is noteworthy and a little difficult.
The legal system as a whole is still uneven. Although no individual right of action has been established, federal legislation passed since 2019 has ordered study on identifying AI-generated content. Publicity rights are governed by a wide range of state legislation. When their voice or likeness is mimicked by AI, performers in states with lax safeguards have few and costly options.
It’s still unknown if the NO FAKES Act will pass the Senate and House in its current form, but it would replace that patchwork with something consistent. Over the last seven years, identical measures have been submitted by Congress several times. There is a sizable stack of deepfake laws that were never put to a vote.
As this develops, it seems that Johansson’s intervention was significant not because of any specific legal power she possessed, but rather because she was prepared to state clearly what the policy argument demanded: the issue is not exclusive to celebrities, it is not limited to antisemitic content, and it will not wait for Congress to find a convenient time.