Insane AI videos of celebs are everywhere —should they embrace it or call their lawyer?

Michael Jackson leans over a KFC table and says, “your chicken’s looking nice, pal,” before swiping the box.

Marilyn Monroe is resurrected as a TSA agent going through traveler’s bags, while the late wheelchair-bound physicist Stephen Hawking is depicted suffering all manner of indignities from dirt biking to being cast as a WWE wrestler.

Clearly this is not how the estates of these people want them portrayed, but new AI video capabilities don’t wait around for permissions and social media is awash with thousands of similar new videos daily.

This has been fueled by the wide release of OpenAI’s Sora 2 app on Oct. 1 along with other video software such as Midjourney and Google Gemini.

Swathes of living actors, actresses, illustrators and animators rushed to their lawyers to protect their likeness. Some, but very few, are embracing it.

Mark Cuban is one of them. “I saw it,” he told The Post of one Sora AI video. “It looks like me hitting the bong hard and going, ‘Oh, s—t. This s—t is good.’”

Was Cuban piqued by the depiction? Not really. “I just posted a comment that said, ‘This is funny. But I don’t smoke.’”

Jake Paul is another, and can be seen converting to Islam, pregnant, putting on make-up and modeling dresses. It should be noted both men are themselves investors in AI.

The capabilities of new text to video apps are phenomenal, able to conjure up practically any scene you can dream up in broadcast quality, from your next-door neighbor landing on the moon or your mother-in-law mining for diamonds.

However, that also means the scope for its abuse remains incredibly high. A lot of the content people produce is puerile, and has come to be labeled “slop.”

OpenAI initially released Sora — which is still currently only available to those who are invited — with a policy of people having to opt-out if they didn’t want to be used in the app.

Following an almighty uproar, in which lawyers rightly pointed out that’s not how US copyright law works, OpenAI reversed course. They have since said they will give “rightsholders more granular control over generation of characters,” which is “similar to the opt-in model for likeness”— although further details are scant.

Aaron Kogan, of Aaron Kogan Management in Los Angeles, hopes eventually there will be a détente between old school movie making and new.

“I share concerns that it will up-end the ecosystem [of Hollywood],” he told The Post. “But I hope that when we find the new equilibrium, there will be a home for AI generated content and there will also be a home for film and television made with people. I just don’t know what that balance will be.”

OpenAI CEO Sam Altman appears in various AI videos — he can been seen driving to the bucket for the Lakers and apparently losing it over getting too little guac at a Chipotle.

Those videos seem relatively cool in contrast to those of the late, famously mellow PBS artist Bob Ross. He’s been depicted getting thrown in jail and painting on the street in his underwear via the magic of AI.

His estate is not happy. Joan Kowalsky, president of Bob Ross, Inc., says his devoted fanbase have been calling to inquire what the heck is going on.

What do they tell Kowalski? “Well,” she says, “for the most part, ‘This is disgraceful. He’s such a wonderful person. Can you do anything about this?’ A lot of people think that because Bob was on public television, he is in the public domain, which is not true.”

Kowalski, who has not personally seen the videos, has been issuing notices. “We’re asking if they will take them down at our request. Or are there other measures we need to take?”

While Ross has a nice guy image, Kowalski makes clear that he was “ferocious about making sure that his name and image were being used properly. He wanted people to be happy with him. If there’s any anxiety coming from that, he would flip.”

Cuban, who has invested in an AI video company called Synthesia, takes the opposite approach.

He can be seen in AI videos coming out as trans, getting arrested or wearing a robe and being tickled.

“I think it’s fun … [and] that what we see now on these AI tools is just the beginning. It’s the first inning of the first preseason game,” he told The Post.

“Like with any new technology, you stand the best chance to leverage it if you start understanding it early on. That’s what I did,” he added.

Cuban is already making it work to his advantage. Any video which uses his likeness also has to flash the name of his discount pharmaceutical company CostPlusDrugs.

However, he opens the tap for people to make videos with his image only sporadically and asks to take down anything that really bugs him: “Someone had me snorting sugar. I just deleted it,” he clarified.

Copyright law is clear cut about how it applies to living people. For those who have passed it can be thornier, and upsetting.

Reacting to videos featuring Robin Williams, who took his own life in 2014 after being diagnosed with Lewy body dementia, his daughter Zelda requested that people “just stop sending me AI videos of Dad … If you’ve got any decency, just stop,” in an Instagram video.

She characterized clips, like Wiliams as a Mexican wrestler and a Walmart greeter, as “gross.”

Mark Roesler — chairman and CEO of CMG Worldwide, which represents the interests of deceased celebrities such as Elvis Presley, James Dean and Albert Einstein — watched a Sora video alongside The Post of Einstein going wild in a fast-food joint and being escorted out by security.

“I’m amazed,” he said after. “Most of my clients would see that and ask, ‘Can you stop it?’ There are certain things we can stop and certain things we can’t” – this, since Einstein is deceased, generally falls into the latter category.

“I don’t like it … But we can’t stop everything … And there are a lot of positive uses that can come out of all this,” he added.

Other Hollywood entities have taken steps to protect their characters. Disney, Universal, and Warner Bros. Discovery, have filed a major legal action against Midjourney, calling the company’s AI generator a “bottomless pit of plagiarism.”

Their lawsuit featured many clear-cut examples where characters including Superman, Batman, Bugs Bunny and Daffy Duck had been replicated by the company’s software.

That lawsuit alleges that to create their characters, Midjourney’s AI must have been illegally trained on them in the first place.

They claim even generic prompts in the engine, such as “classic comic book superhero battle” lead it to create copies of its characters. They’re asking for $150,000 per infringement, which could total billions of dollars.

The company have defended themselves, saying its AI tool was “trained on billions of publicly available images.”

Midjourney claim the responsibility lies with its users to follow its terms of service, which prevent creating works which infringe copyrights.

CEO David Holz used an argument typical of many in AI during a 2022 interview.

“Can a person look at somebody else’s picture and learn from it and make a similar picture? Obviously, it’s allowed for people and if it wasn’t, then it would destroy the whole professional art industry … AIs are learning like people, it’s sort of the same thing and if the images come out differently then it seems like it’s fine,” he told the Associated Press.

The lawsuit is being watched carefully and its results will have a big effect on the future of AI. However, for many, they feel the cat is already out of the bag, so for time being we’d better get used to watching Martin Luther King DJing, Freddie Mercury making video calls and Elvis mowing the lawn.