Illustrations 269318562 © Sakkmesterke | Dreamstime.com
These pictures kiss and tell a damning story. One catch: they’re fake.
A political satire artist says he was banned by artificial intelligence art generator Midjourney for creating a series of fake photos depicting politicians like Joe Biden, Donald Trump, Barack Obama, Bernie Sanders, and Mike Pence cheating on their partners.
Looking at these visuals may send chills down your spine, but Justin T. Brown, who’s also a video editor, says they’re supposed to trigger a reaction. Importantly, he notes that they’re warnings about AI’s potential dangers when put in the wrong hands.
Brown tells PetaPixel that the images, which each took about an hour to create, had been floating around for months, but they only got flagged and taken down two days after they emerged on the Midjourney subreddit.
The artist says the works are part of his series, AI will revolutionize the blackmail industry, to raise awareness about AI-empowered disinformation.
“When used with intelligence and intent, AI can be [a] weapon,” Brown expresses in a Twitter thread chronicling the ban.
“The right AI images or video at the right time could tank the market, cause a riot, or send someone into a pizza shop with a gun.”
Brown tells the publication that for people to realize how damaging AI-generated fake news can be, the images “needed to be a threat,” so he zeroed in on men in power and placed them in compromising scenarios.
“Honestly, these images don’t even have to be THAT high-quality to be believable,” the artist tweets. Some of the fake shots generated feature influential figures in blurry scenarios.
Brown reiterates with the news outlet that he’d “rather not be banned,” but he acknowledges some accountability should be held by users who proliferate content with malicious intent. Still, he asserts that his images were only removed after months, which would have been “too late.”
The synthetic “photos” and the events that have followed reaffirm the need for an international conversation around the largely unregulated world of AI, he notes.
And the conversation’s just starting—better late than never, some may argue. The Vatican (yes, the home of the Pope) has just published a handbook on AI ethics. The UK is also reportedly rolling out a law to have all artificially-generated photos and videos labeled. China has prohibited deepfaked content featuring non-consenting subjects.
[via The Messenger and PetaPixel, images via various sources]