OpenAI is clamping down on deepfake videos of Martin Luther King Jr. on its video tool Sora 2 after his family complained about “disrespectful depictions” of the iconic leader.
“Some users generated disrespectful depictions of Dr. King’s image,” the company said Thursday in a joint statement with the King Estate on social media. “So at King, Inc.’s request, OpenAI has paused generations depicting Dr. King as it strengthens guardrails for historical figures.”
The company said “strong free speech interests” are a concern, but ultimately, public figures and their family should have control over the use of their likenesses.
It added that representatives and estate owners can contact OpenAI to request their images not be used in AI-generated videos.
Bernice A. King, the late civil rights leader’s youngest daughter, contacted OpenAI about taking down the deepfake videos, according to the statement.
Sora 2 – a text-to-video AI app created by Sam Altman’s OpenAI – launched late last month.
The app’s capacity to instantly create lifelike videos has sparked backlash from critics who warn there aren’t enough safety barriers in place.
The app has been used to create mocking, cruel videos of deceased celebrities long after their deaths.
Some clips show physicist Stephen Hawking, who died in 2018, being hoisted by a forklift into the air and then knocked to the ground by WWE wrestlers.
Others depict Elvis Presley, who died in 1977, stumbling and collapsing off the stage in a fake video of his final performance.
The late comedian Robin Williams’ daughter, Zelda Williams, has spoken out about “disturbing” images, audio and video clips made of her deceased father’s likeness with AI.
“I’ve witnessed for YEARS how many people want to train these models to create/recreate actors who cannot consent, like Dad. This isn’t theoretical, it is very very real,” she wrote in an Instagram post earlier this month.
“I’ve already heard AI used to get his ‘voice’ to say whatever people want and while I find it personally disturbing, the ramifications go far beyond my own feelings. Living actors deserve a chance to create characters with their choices, to voice cartoons, to put their HUMAN effort and time into the pursuit of performance.”
Hollywood unions and talent agencies have taken aim at OpenAI over its lack of safety guardrails for actors, as well as its creation of a so-called AI actress, Tilly Norwood.
“The world must be reminded that what moves us isn’t synthetic. It’s human,” SAG-AFTRA President Sean Astin and National Executive Director Duncan Crabtree-Ireland said in a joint statement this month.
They slammed tech firms for creating “a sensationalized narrative, designed to manipulate the public and make space for continued exploitation.”
Actor Scarlett Johansson has accused OpenAI of releasing an AI chatbot with a voice that sounded “eerily similar to mine” – even after she declined to license her voice for a virtual assistant.
Over the weekend, Tom Hanks warned there is an AI-generated video of his likeness circulating promoting “some dental plan” that he has “nothing to do with.”
