The company admitted that Sora had produced “disrespectful” content depicting the civil rights leader. Sora, which has recently gone viral in the United States for its ability to generate hyper-realistic videos, has sparked controversy as users began sharing fake clips of deceased celebrities and historical figures in inappropriate or offensive contexts.
OpenAI stated it would pause all depictions of Dr. King while it works to strengthen its safeguards around the portrayal of historical figures. However, the company continues to permit the creation of videos featuring other prominent individuals — a decision that has drawn criticism.
Clips featuring figures such as President John F. Kennedy, Queen Elizabeth II, and Professor Stephen Hawking have circulated widely on social media. The trend has prompted public outcry from relatives of those portrayed. Zelda Williams, daughter of late actor Robin Williams, urged people to stop sending her AI-generated videos of her father, while Bernice A. King, daughter of Dr. King, echoed that plea: “I concur concerning my father. Please stop.”
Some of the AI-generated videos altered Dr. King’s iconic “I Have a Dream” speech, including one offensive clip reportedly showing him making racist noises. Others depicted lifelike renderings of Dr. King and Malcolm X fighting — further inflaming criticism.
AI ethicist Olivia Gambelin told the BBC that OpenAI’s move was “a good step forward,” but criticized the company for failing to implement such restrictions from the outset. She argued that the use of deepfakes featuring historical figures shows not only a lack of respect but also a risk to public understanding of truth and history.
“It plays too closely with trying to rewrite aspects of history,” she said.
The growing use of deepfakes — videos altered through AI or other digital tools to make people appear to say or do things they never did — has raised widespread concerns about misinformation, discrimination, and digital abuse.
In response, OpenAI said that while it recognized “strong free speech interests in depicting historical figures,” it also believed that these individuals and their families should have control over their likenesses.
“Authorized representatives or estate owners can request that their likeness not be used in Sora cameos,” the company stated.
On Sora, “cameos” allow living people to consent to having their face or likeness used in future AI-generated content.
OpenAI told the BBC earlier in October that it had built “multiple layers of protection to prevent misuse” and was in direct dialogue with public figures and content owners to determine what controls they preferred, with the goal of implementing those measures in future updates.