OpenAI mentioned Monday it might do extra to cease customers of its AI video era app Sora from creating clips with the likenesses of actors and different celebrities after actor Bryan Cranston and the union representing movie and TV actors raised issues that deepfake movies had been being made with out the performers’ consent.
Actor Bryan Cranston, the Display Actors Guild-American Federation of Tv and Radio Artists (SAG-AFTRA) and several other expertise businesses mentioned they struck a take care of the ChatGPT maker over using celebrities’ likenesses in Sora. The joint statement highlights the extraordinary battle between AI firms and rights holders like celebrities’ estates, film studios and expertise businesses — and the way generative AI tech continues to erode actuality for all of us.
Sora, a brand new sister app to ChatGPT, lets customers create and share AI-generated videos. It launched to a lot fanfare three weeks in the past, with AI fans looking for invite codes. However Sora is exclusive amongst AI video turbines and social media apps; it enables you to use different individuals’s recorded likenesses to put them in almost any AI video. It has been, at finest, bizarre and humorous, and at worst, a never-ending scroll of deepfakes which can be almost indistinguishable from actuality.
Cranston seen his likeness was being utilized by Sora customers when the app launched, and the Breaking Unhealthy actor alerted his union. The brand new settlement with the actors’ union and expertise businesses reiterates that celebrities should decide in to having their likenesses accessible to be positioned into AI-generated video. OpenAI mentioned within the assertion that it has “strengthened the guardrails round replication of voice and likeness” and “expressed remorse for these unintentional generations.”
OpenAI does have guardrails in place to forestall the creation of movies of well-known individuals: It rejected my immediate asking for a video of Taylor Swift on stage, for instance. However these guardrails aren’t good, as we have noticed final week with a rising pattern of individuals creating movies that includes Rev. Martin Luther King Jr. They ranged from bizarre deepfakes of the civil rights chief rapping and wrestling within the WWE to overtly racist content material.
Do not miss any of our unbiased tech content material and lab-based evaluations. Add CNET as a most well-liked Google supply.
The flood of “disrespectful depictions,” as OpenAI referred to as them in a statement on Friday, is a part of why the corporate paused the flexibility to create movies that includes King.
Assertion from OpenAI and King Property, Inc.
The Property of Martin Luther King, Jr., Inc. (King, Inc.) and OpenAI have labored collectively to handle how Dr. Martin Luther King Jr.’s likeness is represented in Sora generations. Some customers generated disrespectful depictions of Dr.…— OpenAI Newsroom (@OpenAINewsroom) October 17, 2025
Bernice A. King, his daughter, final week publicly requested individuals to cease sending her AI-generated movies of her father. She was echoing comic Robin Williams’ daughter, Zelda, who referred to as these types of AI movies “gross.”
I concur regarding my father.
Please cease. #RobinWilliams #MLK #AI https://t.co/SImVIP30iN— Be A King (@BerniceKing) October 7, 2025
OpenAI mentioned it “believes public figures and their households ought to finally have management over how their likeness is used” and that “licensed representatives” of public figures and their estates can request that their likeness not be included in Sora. On this case, King’s property is the entity chargeable for selecting how his likeness is used.
This is not the primary time OpenAI has leaned on others to make these calls. Earlier than Sora’s launch, the corporate reportedly instructed a variety of Hollywood-adjacent expertise businesses that they must decide out of getting their mental property included in Sora. However that preliminary method did not sq. with many years of copyright regulation — often, firms must license protected content material earlier than utilizing it — and OpenAI reversed its stance just a few days later. It is one instance of how AI firms and creators are clashing over copyright, together with by high-profile lawsuits.
(Disclosure: Ziff Davis, CNET’s dad or mum firm, in April filed a lawsuit towards OpenAI, alleging it infringed Ziff Davis copyrights in coaching and working its AI techniques.)