Andrew Cuomo isn’t known as the most tech-forward candidate for mayor, but he put artificial intelligence to novel use in his latest ad, which features an AI-generated Cuomo working as a train operator, stock trader and stagehand.
The ad sparked some criticism, including for what it says about human labor.
“It seemed like many of those shots could have been done with a real film crew,” said Manhattan Assemblyman Alex Bores, who has spearheaded legislation regulating AI-generated imagery in politics.
What You Need To Know
- A recent campaign ad features an AI version of Andrew Cuomo working as a subway motorman, stock broker and stagehand
- The ad sparked criticism and raised questions about compliance with a state law that sets strict disclosure rules
- City Council candidate Jonathan Rinaldi, who posted doctored news articles falsely stating he was endorsed by Cuomo and others, defended his posts as “memes”
That imagery is becoming pervasive in New York politics. Before he dropped his reelection bid, Mayor Eric Adams regularly used AI-generated images to attack Democratic nominee Zohran Mamdani and Cuomo.
Incidentally, an AI version of Adams stars in an AI-generated trailer for a new memoir by his ex-girlfriend Jasmine Ray.
A state law passed by Assemblyman Bores requires a disclaimer on AI-generated political content, though there is an exemption for satire or parody.
“The obvious uses of AI — which I think are most of the campaign context — you instantly know are AI, so it’s almost a stylistic choice,” Bores said in an interview. “What worries me a lot more are the ones that are photorealistic, where you have a question as to whether that’s AI.”
Upstate Republican Congresswoman Elise Stefanik furiously objected to an AI video released in July mocking her gubernatorial ambitions. It was created by Democratic challenger Blake Gendebien, who ends the ad with a disclaimer that it “was generated in whole or substantially by artificial intelligence.”
Mayor Adams’ AI content is tagged “AI GENERATED” in the lower right-hand corner. And the Cuomo ad, too, discloses that it “was created with the assistance of artificial intelligence.”
But that may not technically satisfy state law, according to state Sen. Kristen Gonzalez, chair of the Internet and Technology Committee, who co-sponsored the legislation on the political use of AI. The bill specifies the language and size of any disclosure.
“I actually think it really was not in compliance with parts of our bill,” she said. “That disclaimer was not legible for digital media, which is part of the requirements. It’s supposed to be no smaller than other texts — it certainly was much smaller than other texts.”
But Cuomo’s ad is unique: The deepfake is of himself, and the content might qualify as parody, Gonzalez said.
Another AI controversy took shape this week in Queens, where Republican City Council candidate Jonathan Rinaldi has taken great liberties with AI, creating dozens of AI-generated images of himself and others — some photorealistic, including an image that purports to show him delivering a speech at a large pro-Israel rally.
He’s also posted doctored news articles falsely stating he was endorsed by Andrew Cuomo and by Queens Councilman Robert Holden.
And he’s created numerous images giving the false impression that his opponent, Councilwoman Lynn Schulman, has endorsed Mamdani.
In an interview with NY1, Rinaldi defended the social media images as art.
“It’s not fake news,” he said. “They’re memes. When any person tries to go up against the establishment, you have to use every tool at your disposal.”
Rinaldi had previously denied posting the images when contacted by the Queens Daily Eagle — which first reported on them — and claimed his accounts were hacked.
Rinaldi has seemingly opened himself up to legal action, since state law allows victims of political deepfakes to seek relief in court.
As for any further regulation, it’s unlikely to come from Washington, where President Donald Trump has aggressively weaponized AI, including controversial videos posted last week mocking Congressman Hakeem Jeffries and Sen. Charles Schumer.
Assemblyman Bores warns that improving technology means more truly deceptive uses of AI are coming.
“AI is the latest version of CGI, and people have used CGI to make movies and to animate things. This is not saying that we can never do it. But I do think that there is a key distinction when what is being presented is meant to appear like it is real but isn’t,” Bores said. “That is a problem that, as a society, we have to solve.”
