The confines of consent and the ethical use of one's likeness have once again come under scrutiny in the constantly evolving world of technology and celebrities. Hollywood actor Scarlett Johansson has raised her voice against OpenAI, the prominent artificial intelligence research laboratory, following the use of a voice in its ChatGPT product that closely resembles hers. The acclaimed actor revealed in a statement that she was approached by OpenAI nine months prior to the product launch to lend her voice to their AI system. However, she declined the offer, citing "personal reasons."
Johansson expressed shock and anger upon discovering the voice option in ChatGPT, which bore a striking resemblance to her own voice. She emphasized that the similarity was so uncanny that even her closest associates and media outlets couldn't discern the difference.
OpenAI Faces Backlash from Scarlett Johansson Over ChatGPT Voice
In response to widespread criticism regarding the flirtatious and feminine nature of the voice, OpenAI swiftly removed the heavily promoted voice option, dubbed "Sky", from ChatGPT. The company had showcased the capabilities of its new ChatGPT-4o model during a highly publicized event, where researchers engaged with the AI assistant, highlighting Sky's personable and responsive characteristics. The comparison drawn between Sky and Johansson's AI companion character in the 2013 film "Her" was immediate and widespread.
Even OpenAI’s CEO, Sam Altman, seemingly hinted at the intentional resemblance to Johansson’s character by posting a cryptic tweet after the presentation simply saying “her”. However, OpenAI later clarified that Sky was not based on Johansson, asserting that it belonged to a different professional actress using her natural speaking voice. The company, however, refrained from disclosing the identity of the voice talent to protect their privacy.
Johansson disclosed that Altman initially approached her with the belief that her voicing the system could bridge the gap between tech companies and creatives, fostering consumer comfort amid the transformative impact of human-AI interaction. Despite declining the offer initially, Altman allegedly reached out to her agent again, just two days before the release of Sky, prompting Johansson's legal team to intervene and demand the removal of the voice.
While some commentators highlighted the similarities between Sky and Johansson's character in Her, others criticised the voice for its fawning and gendered characteristics. The Daily Show host, Desi Lydic, quipped, "You can really tell that a man built this tech. She’s like, ‘I have all the information in the world, but I don’t know anything.’
OpenAI defended its selection of ChatGPT’s vocals, asserting that they were chosen based on criteria such as timelessness and an approachable quality that inspires trust. The company claimed to have reviewed hundreds of voice-acting submissions over five months before selecting the final options. However, the controversy surrounding Sky intensified in the wake of several top members of OpenAI’s safety team resigning, citing prioritisation of product innovation over safety protocols.
The Unauthorised AI Advertisement
In another instance, Johansson recently filed a lawsuit against an AI-driven advertisement that exploited her name and image without her consent, sparking a heated debate over the nexus of artificial intelligence and personal rights. The crux of the controversy revolves around a 22-second ad that was disseminated on X, the platform formerly known as Twitter. This advertisement was crafted by a peculiar AI-driven app known as Lisa AI: 90’s Yearbook & Avatar. What made it even more audacious was its unapologetic use of real footage of Scarlett Johansson, which was manipulated to generate a fictitious image and dialogue, all without her consent.
The video commenced with an exclusive behind-the-scenes clip of Johansson from her role in the Marvel film "Black Widow." With a welcoming "What’s up, guys? It’s Scarlett, and I want you to come with me," she beckoned to her audience before the screen transitioned seamlessly into AI-generated photos that strikingly resembled her. A voice, skillfully mimicking Scarlett Johansson's, continued to promote the app with a persuasive tone: "It's not limited to avatars only. You can also create images with texts and even your AI videos. I think you shouldn’t miss it," the imitation concluded.
Legal Action and Consequences
Her representatives were swift to clarify this in a statement to Variety, emphasising that she had taken immediate legal action as soon as the unauthorized advertisement came to her attention. Her lawyer, Kevin Yorn, stated, "We do not take these things lightly. Per our usual course of action in these circumstances, we will deal with it with all legal remedies that we will have." As a consequence, the advertisement was promptly removed from the platform.
However, the fine print under the advertisement succinctly noted, "Images produced by Lisa AI. It has nothing to do with this person." This raises questions about the responsibilities of both the platform and the app developer in safeguarding the rights and identities of individuals.
A Wider Issue
This isn't the first time Scarlett Johansson has been confronted with the unauthorized use of her image. In 2018, she vocalised her distress in an interview with the Washington Post, shedding light on the sinister world of "deepfakes." These are computer-generated simulations that illicitly superimpose women's faces onto pornographic videos.
Johansson's lamentation still rings true today: "Nothing can stop someone from cutting and pasting my image or anyone else's onto a different body and making it look as eerily realistic as desired." The virtual landscape remains largely unregulated, a digital wild west that remains predominantly lawless, even in the face of policies that apply only to a limited jurisdiction.
Johansson is not alone in her struggle to protect her image from unauthorized AI use. In a similar vein, Tom Hanks recently took to Instagram to alert his fans about an AI-generated image being used in a dental plan promotional video without his consent. His message was, "Beware! … I have nothing to do with it," highlights the broader concern of celebrities having their identities manipulated without consent.
Kartik Aaryan, a Bollywood actor, recently faced a distressing incident. A morphed video of him endorsing the Congress party in the upcoming Madhya Pradesh elections went viral. However, the video's origin was far from political; it was originally shared by the OTT platform Disney+Hotstar for the ICC Men's World Cup. The video was cleverly edited and dubbed to portray the actor supporting the political party's campaign promises. The video not only exploited Kartik Aaryan's image but also propagated a false narrative. This incident highlights how AI-generated content can be manipulated for personal and political agendas.
Furthermore, authors, including the well-known comedian Sarah Silverman, have filed lawsuits against ChatGPT, developed by OpenAI and Meta, for copyright infringement. Their allegations suggest that artificial intelligence models have been trained on their work without proper authorization, prompting a larger discussion about intellectual property and the use of AI in creative endeavours.
The Power of Influence and the Challenge of Verification
In today's digital age, the power of influence is amplified by the widespread reach of social media. Celebrities, with their vast followings, can shape public opinion and create trends. However, this power also makes them vulnerable to exploitation. Their images and statements can be taken out of context, manipulated, or entirely fabricated to serve personal and political agendas.
This challenge is not limited to celebrities alone. Common people, with the advent of deepfake technology, can also be targeted. These maliciously crafted AI-generated images and videos can tarnish reputations, influence opinions, and even incite social or political unrest.
The cases of Scarlett Johansson, Tom Hanks, Kartik Aaryan, and others emphasize the need for stringent regulations and ethical considerations in the use of AI-generated images. While technology offers a plethora of creative possibilities, it also raises ethical, legal, and privacy challenges that demand thoughtful solutions.
Views expressed by the author are their own.
Suggested reading: Rise Of AI Girlfriends In The US Leads To Epidemic, 'Male Loneliness'
/shethepeople/media/agency_attachments/iUYfaBbcMozaYLoGnoov.png)
/shethepeople/media/member_avatars/h4HZwmcBacu03I80QTAh.jpg )
Follow Us/shethepeople/media/media_files/EEnyXSDVBArwDVXY5cn7.jpg)