OpenAI has released Sora, a new state-of-the-art AI model with deepfake powers. Most particularly, it can create extremely convincing fakes of celebrities and historical figures. While launched with a spirit of creative potential, Sora nonetheless poses troubling ethical issues around the attempt to resurrect and exploit the likenesses of deceased persons. The application has sparked discussions around intellectual property rights, particularly as it relates to the legacy of those who can no longer voice their opinions.
Sora is able to produce videos with well known figures such as Robin Williams and Martin Luther King Jr. It has been used to generate dialogue between illustrative fictional characters, such as Family Guy’s Peter Griffin and Pokémon’s Pikachu. Once Sora was released, it quickly became a magnet for a sea of AI-generated junk that relied on copyrighted characters. This new wave raised alarm for many content creators and IP owners with concerns of unlawful utilizations.
OpenAI CEO Sam Altman has been unequivocal about this. He explained that for Hollywood studios and agencies, they can choose to opt-out if they want to keep their intellectual property out of content produced by Sora. Then, within a week, he declared that Sora would introduce granular opt-in copyright controls. This change is intended to give creators clearer, more granular permissions to be able to use people’s likenesses. Advocates are quick to point out it does an inadequate job of defending the legacies of those no longer living.
Ethical questions surrounding the use of Sora to generate deepfakes of deceased public figures have been raised. Critics argue that OpenAI has taken a “shoot-first, ask-questions-later” stance on copyright law. The ability to generate videos of figures like Franklin Delano Roosevelt and Richard Nixon raises questions about respect for their memory and the authenticity of their messages.
Robin Williams’ daughter Zelda Williams took to Twitter last night to express her outrage and sadness. She’s been opposed to the use of Sora to make animated videos starring her father. Most notably, she has implored her fans not to send her AI-generated video clips. Yet she stressed that these portrayals make her feel no comfort.
“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
Sora’s policy does not allow for the production of videos with people who have died recently. This encompasses some heavyweights such as former President Jimmy Carter and Michael Jackson. This decision addresses the delicacy of depicting subjects who have only recently passed away. Most importantly, it ignites broader discussions regarding the moral ramifications associated with deepfake technology.
As the tech becomes increasingly realistic, doubts about how easily it can be misused are only heightened. As critics pointed out, producing videos that just “vaguely resemble and sound like” those who have passed reduces their lives and legacies to distractions. Robin Williams himself voiced frustration over this issue, stating:
“To watch the legacies of real people be condensed down to ‘this vaguely looks and sounds like them so that’s enough,’ just so other people can churn out horrible TikTok slop puppeteering them is maddening.” – Robin Williams
The continued discussion around Sora sheds light on the difficult challenges at play when trying to prioritize technological advancement with ethical awareness. As the AI landscape continues to change, we need to get ahead of the curve and define how deepfake technology should be used. This is particularly important for people who are no longer able to speak up for themselves.
OpenAI has not returned inquiries about whether deepfaking deceased individuals is permissible under the created content guidelines. This raises many obvious questions. The voice for this issue is still loud, as advocates on all sides continue to debate the meaning and impact of what Sora can do.
