Authentic Monet Painting Labeled "Made by AI" Fooled Thousands of Online Critics
In May, an anonymous conceptual artist turned social media into an unwitting psychological laboratory: they posted a genuine Claude Monet painting on X with a fake "Made with AI" label—and watched as users fell over themselves to criticize it as a soulless, poorly composed smear.
The experiment was conducted by a pseudonymous digital artist known as SHL0MS. On May 12, they posted a cropped image from Monet's famous Water Lilies series—a cycle of about 250 oil paintings that the French Impressionist created over the last three decades of his life—with a caption inviting users to explain "why this is worse than a real Monet."
The reactions were swift—confident and categorical. "No cohesion in the choice of depth and color," wrote one user. "The reflection of the tree blends into the water lilies without the slightest understanding of spatial depth and contrast." Another declared, "Monet actually understood how light behaves on water." Some dismissed the work as "soulless," while others outright called it "garbage" and demanded that "AI be kept as far away as possible."
One particularly zealous critic wrote an 850-word breakdown, detailing all the flaws of the supposed neural network artwork. Others drew "line of sight" diagrams to visually demonstrate what they considered an incoherent composition. When the post went viral and the truth was revealed, many began quietly deleting their comments—but not before SHL0MS and other users managed to take screenshots.
This experiment echoes the results of peer-reviewed research. A 2024 study published in the journal Nature found that participants generally preferred AI-generated artwork over human-made pieces when they didn't know their origin; however, once they were told the author was a machine, their attitude toward the works shifted for the worse. An earlier 2004 study on the "effort heuristic" similarly showed that people value art more highly when they believe a significant amount of human effort went into its creation.
This experiment has reignited a large-scale debate about what some commentators are calling an AI "witch hunt" within online art communities. Digital artists are reporting false accusations of using generative AI, and some are feeling forced to abandon experimental techniques to avoid raising suspicion. One digital artist wrote in March 2026 that their sketches from years ago—created long before the advent of tools like Midjourney and DALL-E—had been flagged as AI-generated.
SHL0MS, who describes themself as an "NFT performance artist," is no stranger to executing provocative conceptual projects: their track record includes selling a blank-image NFT and blowing up a Lamborghini to auction off its fragments. The Monet experiment fits seamlessly into this tradition: the artist uses the internet's knee-jerk reactions to expose the mechanisms behind how human judgments are formed.
As one X user aptly summarized: "A clear demonstration of how a single label changes what people see."
