OpenAI announced Thursday that it has suspended generating videos impersonating the late civil rights activist Martin Luther King Jr. through its artificial intelligence tool, Sora. This decision follows a specific request from Dr. King’s estate, after users produced “disrespectful depictions” of his image.
The company justifies this measure by the respect owed to public figures and their loved ones, stating that the latter should have the final say over the use of their image. OpenAI specifies that rights holders or official representatives can now request “the removal of any representation generated by Sora.”
This restriction comes shortly after the launch of Sora, a platform that allows the creation of realistic videos with generative AI, and raises questions about the risks associated with this technology. Controversial videos, such as those showing King in degrading situations or fictional confrontations with Malcolm X, circulated quickly, fueling fears about the manipulation of images of historical figures.
Bernice King, daughter of the activist, publicly called for an end to these practices, joining the protests of other families, such as that of Robin Williams. These incidents underline the urgency of regulating the use of artificial intelligence tools “to preserve the dignity of deceased personalities.”
Broader legal and cultural issues
Beyond ethical issues, Sora also raises legal challenges, particularly regarding copyrighted works. The platform is full of content generated from protected characters or universes, such as “SpongeBob SquarePants” or “South Park”, without the rights holders having been consulted.
States are stepping forward to defend “intellectual property.” According to the Japanese media IT Media, the Japanese government has officially asked OpenAI to stop copying its anime on the Sora video platform. At a press conference on October 10, Minoru Kiuchi, Minister of Economic Security and Intellectual Property, alleged copyright infringement.
He recalled that “manga and anime are irreplaceable cultural treasures” for Japan. Given the rapid spread of videos that imitate famous works, the government threatens to use the law on the promotion of AI to regulate these practices.
For his part, Sam Altman, CEO of OpenAI, said in a blog post that it offers “more precise control over character generation, similar to the similarity membership model, but with additional controls.” Faced with growing pressure on these issues, the American company appears to be adopting a more cautious moderation strategy with its video generation tool, in contrast to its recent decision to authorize… “erotic” conversations on ChatGPT.
Source: BFM TV
