Introduction to AI-Generated Videos
In a parallel reality, the Queen Elizabeth II marvels at cheese balls, a heavily armed Saddam Hussein struts in a wrestling ring, and Pope John Paul II attempts to skateboard. Hyperrealistic videos of deceased celebrities, created using user-friendly AI applications like Sora from OpenAI, are flooding social media platforms and sparking a debate about the control of deceased individuals’ images.
OpenAI’s Sora Application
Launched in September, OpenAI’s Sora is widely regarded as a “deepfake” machine, generating hyperrealistic yet false content through AI. This application has produced a flood of videos featuring historical figures such as Winston Churchill, alongside celebrities like Michael Jackson and Elvis Presley.
Controversial Videos and Reactions
In a TikTok video reviewed by AFP, Queen Elizabeth II, adorned with a pearl necklace and crown, arrives at a wrestling match on a scooter, crosses the rope, and jumps over a wrestler. Other clips on Facebook show her praising cheese balls in a supermarket aisle, while another depicts her playing soccer.
However, not all videos generated by OpenAI’s Sora model have elicited laughter. In October, the AI giant prevented users from creating videos using Martin Luther King Jr.’s image due to complaints from his heirs. Some users had created videos where this civil rights movement icon made monkey noises during his famous “I Have a Dream” speech and uttered offensive, crude, or racist comments.
“Unnerving” Concerns
Constance de Saint Laurent, a professor at the University of Maynooth in Ireland, explained to AFP that we are entering the “uncanny valley,” a theory suggesting that when an object reaches a certain level of anthropomorphic likeness or hyperrealism, it evokes feelings of unease and discomfort.
“If you suddenly started receiving videos of a loved one who has passed away, it would be traumatic,” she said. “These videos have real-life consequences.”
In recent weeks, the children of actor Robin Williams and U.S. activist Malcolm X have condemned the use of Sora to create videos of their deceased parents.
Zelda Williams, Robin Williams’ daughter, recently requested on Instagram that people stop sending her AI-generated clips of her father. She called it “unnerving.”
An OpenAI representative told AFP that while “freedom of expression strongly supports the representation of historical figures,” public personalities and their families should ultimately control their images.
For “recently deceased” figures, authorized representatives can now request that their images not be used in Sora, the representative added.
Erosion of Trust
Despite OpenAI’s claims of allowing people to control their images, Hany Farid, co-founder of U.S. cybersecurity firm GetReal Security and a professor at the University of California, Berkeley, pointed out to AFP that they have published a tool that contradicts this intention.
“Although (largely) they have ended the creation of Martin Luther King Jr. videos, they do not prevent users from appropriating the identities of many other celebrities,” Farid lamented.
“Beyond the protections OpenAI implements for that particular figure, another AI model will not have these safeguards, and thus this problem will only worsen,” he warned.
This became evident after the recent murder of Hollywood director Rob Reiner, when AFP fact-checkers discovered AI-generated clips of his image circulating online.
Broader Implications and Concerns
As advanced AI tools proliferate, vulnerability is no longer limited to public figures; even non-famous deceased individuals can have their names, images, and words distorted for manipulation purposes.
Researchers warn that the uncontrolled spread of synthetic content, known as “AI slop,” could eventually drive users away from social media.
“The issue of misinformation in general isn’t that people believe it. Many don’t,” Saint Laurent said. “The problem is that they see genuine information and don’t trust it to be true, and tools like Sora will ‘massively amplify this phenomenon.'”
Key Questions and Answers
- What is OpenAI’s Sora application? Sora is an AI-powered tool developed by OpenAI that allows users to create hyperrealistic videos using their input. It has been criticized for enabling the creation of misleading content featuring deceased celebrities and historical figures.
- Why are people concerned about AI-generated videos of deceased celebrities? The uncanny valley effect, trauma from receiving videos of deceased loved ones, and the potential for misinformation and manipulation are among the concerns.
- How has OpenAI responded to criticism? OpenAI has stated that public personalities and their families should have control over their images, allowing authorized representatives to request the removal of deceased individuals’ likenesses from generated content.
- What are the broader implications of AI-generated content? As AI tools become more sophisticated, the risk of misinformation and manipulation extends beyond public figures to include non-famous deceased individuals, potentially eroding trust in online platforms.
