OpenAI’s new AI model, Sora, has stirred considerable controversy about the ethics of deepfaking dead people. This powerful new technology allows users to generate videos of people that looks incredibly real, from celebrities to long-dead political figures. It has made us consider the vital questions regarding creativity and honoring the dead. The model has already proven capable of generating video of historical figures such as Robin Williams and Martin Luther King, Jr. It has included historical leaders, like Franklin Delano Roosevelt.
Sora’s expanded capabilities make it possible to produce deepfakes of both iconic celebrities and well-known historical figures. The public roll out of this technology has incensed families and fans. Zelda Williams, daughter of the late actor Robin Williams, has publicly condemned the use of AI-generated videos that depict her father. She’s called on her fans to refrain from sending her that kind of material. Most importantly, it profoundly upsets her to witness the way her father is exhibited via these platforms.
“Please, just stop sending me AI videos of Dad. Stop believing I wanna see it or that I’ll understand. I don’t and I won’t,” – Zelda Williams
The conversation surrounding Sora is multifaceted. Sam Altman, the CEO of OpenAI, did not mince words. He stated that Hollywood studios and agencies need to proactively opt-out if they don’t want to see their intellectual property in Sora-generated videos. This statement anecdotally reflected apprehension regarding the issues of consent and data ownership. Yet Altman soon announced that Sora would provide a more granular copyright opt-in with custom provisions as the better way to tackle these problems.
The consequences of using Sora go farther than just something fun. The model has produced deepfake replicas of other deceased celebrities like Bob Ross, John Lennon, and Alex Trebek. This troubling circumstance raises urgent ethical dilemmas. Is it ethical to use the images of people who cannot give consent to being created again? Unfortunately, Sora is not programmed to produce original work on people who have just passed away. This list includes unusual bedfellows—people like former President Jimmy Carter and pop icon Michael Jackson.
Many critics contend that the technology commercializes and trivializes the legacies of actual human beings. Zelda Williams shared her feelings on this matter, stating:
“If you’ve got any decency, just stop doing this to him and to me, to everyone even, full stop. It’s dumb, it’s a waste of time and energy, and believe me, it’s NOT what he’d want.”
Unfortunately, the arrival of Sora has ignited a new flood of AI-generated videos featuring otherwise copyrighted characters. This onslaught brings with it troubling threats to intellectual property protections rights and dangers of censorship. Countless users have gamed the model to create videos starring characters such as Peter Griffin and Pikachu without appropriate rights.
Furthermore, OpenAI has been nonresponsive on the permissibility of deepfaking dead people. Our repeated requests for statements addressing the ethical implications of this practice have been met with silence from the organization.
As technology advances, the creativity vs. disrespect debate will become less black and white. A CGI character like Sora is capable of reproducing human likenesses with unbelievable accuracy. This in turn raises challenging questions regarding ownership and morality in the digital content creation space. As much as that demand increases the opportunity for innovators’ new storytelling tools, it increases the responsibility that comes with them.