A union representing workers in the creative industries says there is “real concern” among its members about a $1bn (£740m) deal which will see Disney characters brought to ChatGPT and video-generation tool Sora.
It is the first major studio to license parts of its catalogue to OpenAI, which makes the tools.
Fans will be able to generate and share pictures and videos of more than 200 characters from Disney’s franchises, including Pixar, Marvel and Star Wars.
“Everyone in the entertainment industry, especially all the creative talent, are incredibly worried about what the implications are,” said Sag-Aftra executive director Duncan Crabtree-Ireland.
He said the two companies had contacted the union – which represents 170,000 media professionals worldwide – to say the agreement specifically rules out the use of the image, likeness or voice of human performers.
But he said it was still a worrying development.
“There’s real concern because nobody wants to see human creativity given away to AI models,” he told BBC News.
The move comes as OpenAI faces mounting questions about how its rapidly advancing tech is used – and as anxiety in Hollywood increases over the impact of AI on the creative industries.
According to a blog post announcing the news, the list of eligible characters include those from Disney films Zootopia, Moana and Encanto – as well as characters like Star Wars’ Luke Skywalker and Marvel’s Deadpool.
It, of course, also includes Mickey and Minnie Mouse.
But quite how the characters will sound remains unclear, as Disney said the agreement “does not include any talent likenesses or voices”.
“The rapid advancement of artificial intelligence marks an important moment for our industry,” Disney boss Bob Iger said.
“Through this collaboration with OpenAI we will thoughtfully and responsibly extend the reach of our storytelling.”
People are expected to be able to begin making the videos and images in Sora and ChatGPT in early 2026.
The deal comes amid reports Disney lawyers sent Google a cease-and-desist letter on Wednesday accusing the tech firm of infringing its copyrights “on a massive scale”.
The BBC has asked Google for a response.
Joel Smith, intellectual property partner at law firm Simmons & Simmons, said Disney’s OpenAI deal showed “rights owners and major AI developers are moving fast to strike collaborative licensing deals to access content for training and future use”.
Equity, the entertainment trade union, says actors’ rights need to be protected. The group is currently balloting thousands of performers asking if they are willing to refuse being digitally scanned on set to push for stronger AI safeguards.
Cathy Sweet, Equity’s head of TV and Film, said the deal between Disney and OpenAI reinforced “exactly why our members are fighting for AI protections”.
“The recorded material that will be sold on as part of this eye-watering big money deal are the result of creatives’ professional work, and their rights must be protected,” she said.
Sora’s hyper-realistic videos have proven to be popular in the US, but there have been significant criticisms from those claiming it has also led to some people creating offensive deepfakes of dead public figures.
In October, OpenAI paused the video tool’s ability to generate images of Dr Martin Luther King Jr after the app produced “disrespectful” depictions of the civil rights leader.
The firm acknowledged the need for stronger safeguards after clips emerged showing him saying offensive things, prompting a public outcry.
Other similar videos of President John F. Kennedy, Queen Elizabeth II and physicist and cosmologist Stephen Hawking also circulated widely.
In some cases, family members called on OpenAI to stop allowing videos to be made.
Zelda Williams, daughter of the late comedian Robin Williams, urged people to stop sending her AI-generated videos of her father.
In November, Warner Music Group announced plans to launch an AI music venture with technology start-up Suno, a year after it sued the firm in a landmark case.
