A fake, AI-generated image of Pope Francis stepping out in a sleek white puffer jacket and a bejeweled crucifix attracted millions of visitors this weekend, with many mistaking it for the real image.
Experts fear that the rapidly advancing technology behind images is rapidly eroding our ability to distinguish fake photos, which can be created in seconds, from reality.
The Midjourney AI program was used to create an image of the Pope, which was apparently first posted on a Reddit page dedicated to AI-generated art and later posted on Twitter, where several viral posts appeared. Reached millions of views. but didn’t say it wasn’t real.
This came days after Eliot Higgins, founder of the investigative journalism group call the catused Midjourney to capture vivid photos of the arrest of former US President Donald Trump.
Although he clarified that they were created by artificial intelligence, other users copied and distributed the images without attribution. Mr. Higgins later said that his developers forbade him from using the AI tool.
Midjourney is an artificial intelligence technology that creates graphics from simple text prompts. While it has been possible to fake images with Photoshop for years, no knowledge is required to use this tool and it only takes a few seconds to create photorealistic fakes. Other similar programs are DALL-E and Stable Diffusion by Open AI.
Henry Eider, AI expert and host of a BBC radio series. The future is synthesizedsaid I that this technology has developed at lightning speed over the past year and shows no signs of slowing down.
Not only are the tools “radically accessible” and easy to use for everyone, they’ve also become more sophisticated and produce images that look more and more realistic.
“A lot of people look really good in this look, especially at first glance, people might think that the Pope is really adopting an Italian fashion sense.”
But Mr Ager said there are still visible flaws that confirm recent viral images are fakes, including images of some body parts.
Examples include the Pope’s right hand trying to hold a water bottle in what looks like a “slippery mess” or Trump’s tousled hair. Images can also have a hyper-realistic “sticky” quality that suggests something is wrong.
However, Mr. Adjer said that these inconsistencies could be ironed out within months or years. People may not distinguish a real picture from a fake in a year or two, even if they look closely.
So far, most of the users who share AI art on Twitter clearly label it and do not intend to mislead them. OpenAI and Midjourney are also working on security restrictions on queries users can enter to generate images, Ajer said. But the trend is worrisome, especially when the technology is used for malicious purposes and falls into the wrong hands.
He added: “Some of these images are already very, very difficult to tell if they are real or not. This gives us an idea of how bad actors, disinformation agents, can use these tools as weapons, especially when some of the errors in the current model start to learn and disappear.
Dr. Mhairi Aitken, Ethics Fellow at the Alan Turing Institute, explained: I that images could be even more risky if applied to ordinary people who don’t have the means to stop them from spreading.
They can be used to create false, compromising images, showing people places they shouldn’t be, or they can be used as a form of bribery or humiliation, or, more broadly, to undermine political processes.
Dr. Aitken said one of the more worrying side effects is that these images can make people question the reality of the real images that appear in traditional media.
She said: “This could be footage of the war in Ukraine, and if challenged as potentially fake, it would change history significantly. It fundamentally changes what people think is true and has implications for political attitudes and democratic processes.”
Even disinformation experts can be vulnerable. Dr. Daniel Jolly, an assistant professor of disinformation at the University of Nottingham, admitted that while browsing Twitter, he did not realize that the Pope’s photo was a fake.
He said: “I really laughed and thought what a weird thing for a dad, I didn’t even know it was fake. The image is of very good quality – and that’s the danger of this kind of misinformation because it looks so good.”
He said that when images match people’s previous beliefs, they are less likely to challenge them because they seem “real” to them, while re-exposure to false images and stories can also convince people of ideas that aren’t true. .
Source: I News
With a background in journalism and a passion for technology, I am an experienced writer and editor. As an author at 24 News Reporter, I specialize in writing about the latest news and developments within the tech industry. My work has been featured on various publications including Wired Magazine and Engadget.