The pope did not wear Balenciaga. And filmmakers did not fake the moon landing.
But artificial intelligence image generators have created photorealistic scenes depicting such images that are capable of fooling the casual viewer.
The technology, including Midjourney and OpenAI’s DALL-E, has delighted a burgeoning crop of artists who specialise in AI fakery. Misinformation watchdogs, however, are growing concerned that the technology will make it harder than ever to determine what is real and fake online.
Enter the AI detectors. A new crop of software is being developed that claims to use its own set of AI algorithms to determine whether an image was crafted by digital hands.
In tests by The New York Times, the software performed well in many cases, often identifying AI images that might fool the naked eye. But the technology was not perfect, missing some obvious images, and experts said it will struggle to keep pace with AI generators that are constantly improving.
One image appears to show billionaire entrepreneur Elon Musk embracing a lifelike robot. The image was created using Midjourney, an AI image generator, by Guerrero Art, an artist who works with AI technology.
Despite the implausibility of the image, it managed to fool several AI-image detectors.
The detectors, including versions that charge for access, such as Sensity, and free ones, such as Umm-maybe’s AI Art Detector, are designed to detect difficult-to-spot markers embedded in AI-generated images. They look for unusual patterns in how the pixels are arranged, including in their sharpness and contrast. Those signals tend to be generated when AI programs create images.
But the detectors ignore all context clues, so they don’t process the existence of a lifelike automaton in a photo with Musk as unlikely. That is one shortcoming of relying on the technology to detect fakes.
Several companies, including Sensity, Hive and Inholo, the company behind Illuminarty, did not dispute the results and said their systems were always improving to keep up with the latest advancements in AI-image generation. Hive, an image-detection tool, added that its misclassifications may result when it analyses lower-quality images. Umm-maybe and Optic, the company behind AI or Not, did not respond to requests for comment.
Detection technology has been heralded as one way to mitigate the harm from AI images.
AI experts such as Chenhao Tan, an assistant professor of computer science at the University of Chicago and the director of its Chicago Human+AI research lab, are less convinced.
“In general, I don’t think they’re great, and I’m not optimistic that they will be,” he said. “In the short term, it is possible that they will be able to perform with some accuracy, but in the long run, anything special a human does with images, AI will be able to re-create as well, and it will be very difficult to distinguish the difference.”
Most of the concern has been on lifelike portraits. Florida Governor Ron DeSantis, a Republican candidate for president, was criticised when his campaign used AI-generated images in a post. Synthetically generated artwork that focuses on scenery has also caused confusion in political races.
Many of the companies behind AI detectors acknowledged that their tools were imperfect and warned of a technological arms race: the detectors must often play catch-up to AI systems that seem to be improving by the minute.
“Every time somebody builds a better generator, people build better discriminators, and then people use the better discriminator to build a better generator,” said Cynthia Rudin, a computer science and engineering professor at Duke University, where she is also the principal investigator at the Interpretable Machine Learning Lab. “The generators are designed to be able to fool a detector.”
Sometimes, the detectors fail even when an image is obviously fake.
Dan Lytle, an artist who works with AI and runs a TikTok account called The_AI_Experiment, asked Midjourney to create a vintage picture of a giant Neanderthal standing among normal men. It produced an aged portrait of a towering, yet-like beast next to a quaint couple.
The wrong result from each service tested demonstrates a drawback with the current AI detectors: they tend to struggle with images that have been altered from their original output or are of low quality, according to Kevin Guo, co-founder and CEO of Hive.
When AI generators such as Midjourney create photorealistic artwork, they pack the image with millions of pixels, each containing clues about its origins. “But if you distort it, if you resize it, lower the resolution, all that stuff, by definition you’re altering those pixels and that additional digital signal is going away,” Guo said.
When Hive, for example, ran a higher-resolution version of the yeti artwork, it correctly determined the image was AI-generated.
Such shortfalls can undermine the potential for AI detectors to become a weapon against fake content. As images go viral online, they are often copied, resaved, shrunken or cropped, obscuring the important signals that AI detectors rely on. A new tool from Adobe Photoshop, known as generative fill, uses AI to expand a photo beyond its borders. (When tested on a photograph that was expanded using generative fill, the technology confused most detection services.)
An unusual portrait of President Joe Biden, taken in Gettysburg, Pennsylvania, by Damon Winter, a photographer for the Times, has much better resolution. Many of the detectors correctly thought the portrait was genuine; but not all did.
Falsely labelling a genuine image as AI-generated is a significant risk with AI detectors. Sensity was able to correctly label most AI images as artificial. But the same tool incorrectly labelled many real photographs as AI-generated.
Those risks could extend to artists, who could be inaccurately accused of using AI tools in creating their artwork.
A Jackson Pollock painting called Convergence features the artist’s familiar, colourful paint splatters. Most — but not all — of the AI detectors determined the image was real and not an AI-generated replica.
Illuminarty’s creators said they wanted a detector capable of identifying fake artwork, such as paintings and drawings.
In the tests, Illuminarty correctly assessed most real photos as authentic but labelled only about half of the AI images as artificial. The tool, creators said, has an intentionally cautious design to avoid falsely accusing artists of using AI.
Illuminarty’s tool, along with most other detectors, correctly identified a similar image in the style of Pollock that was created by the Times using Midjourney.
AI-detection companies say their services are designed to help promote transparency and accountability, helping to flag misinformation, fraud, nonconsensual pornography, artistic dishonesty and other abuses of the technology. Industry experts warn that financial markets and voters could become vulnerable to AI trickery.
One image, in the style of a black-and-white portrait, is fairly convincing. It was created with Midjourney by Marc Fibbens, a New Zealand-based artist who works with AI. Most of the AI detectors still managed to correctly identify it as fake.
Yet, the AI detectors struggled after just a bit of grain was introduced. Detectors such as Hive suddenly believed the fake images were real photos.
The subtle texture, which was nearly invisible to the naked eye, interfered with its ability to analyse the pixels for signs of AI-generated content. Some companies are now trying to identify the use of AI in images by evaluating perspective or the size of subjects’ limbs, in addition to scrutinising pixels.
AI is capable of generating more than realistic images — the technology is already creating text, audio and videos that have fooled professors, scammed consumers and been used in attempts to turn the tide of war.
AI-detection tools should not be the only defence, researchers said. Image creators should embed watermarks into their work, said S. Shyam Sundar, the director of the Center for Socially Responsible Artificial Intelligence at Pennsylvania State University. Websites could incorporate detection tools into their backends, he said, so that they can automatically identify AI images and serve them more carefully to users with warnings and limitations on how they are shared.
Images are especially powerful, Sundar said, because they “have that tendency to cause a visceral response. People are much more likely to believe their eyes.”
This article originally appeared in The New York Times.
Written by: Stuart A. Thompson and Tiffany Hsu
©2023 THE NEW YORK TIMES