Here's a concerning dispatch from the fresh pits of AI-assisted misinformation hell.
A fake image of an explosion at the Pentagon, almost certainly generated by an AI, garnered so much attention on Twitter this morning that, according to Insider, it caused the stock market to briefly dip.
The picture was shared by a "verified" account — though of course that means nothing now that you can just pay for a checkmark, with no actual verification process — called "Bloomberg Feed" on Twitter with the misleading caption "Large Explosion Near the Pentagon Complex in Washington, DC — Initial Report."
The post circulated widely enough on the platform to have a real-world impact. After Twitter user DeItaone, who has over 650,000 followers, shared the post at 10:06 am, the stock market fell 0.26 percent a mere four minutes later, according to Insider.
Though the market recovered pretty quickly, the ordeal is a great example of just how quickly and effectively AI-generated misinformation can move through our existing information pathways — especially on a platform as deeply flawed as Elon Musk's Twitter.
Prime example of the dangers in the pay-to-verify system: This account, which tweeted a (very likely AI-generated) photo of a (fake) story about an explosion at the Pentagon, looks at first glance like a legit Bloomberg news feed. pic.twitter.com/SThErCln0p
— Andy Campbell (@AndyBCampbell) May 22, 2023
Definitely Not Real
While we're still not 100 percent sure the somewhat convincing-looking image was generated by an AI, it certainly has some of the typical hallmarks of one. For instance, the fence around the building appears to melt into the sidewalk, while the building's window frames aren't perfectly aligned.
It's also worth noting that the market responded negatively even after law enforcement agencies took to Twitter to debunk the image.
"[The Pentagon Force Protection Agency] and the [Arlington County Fire Department] are aware of a social media report circulating online about an explosion near the Pentagon," Arlington's fire department tweeted over half an hour before DeIatone retweeted the image. "There is NO explosion or incident taking place at or near the Pentagon reservation, and there is no immediate danger or hazards to the public."
Twitter has also replaced the original post with a disclaimer.
"This tweet is based on an AI-generated hoax," the disclaimer reads. "The initial report itself was fraudulent and later deleted."
Nonetheless, plenty of other accounts chose to reshare the image, including several conspiracy-affiliated accounts, as well as Russian state media account RT, which alone has over three million followers.
The incident goes to show that the use of widely available generative AI tools can have very real consequences. And considering that we're only beginning to scratch the surface of what's possible with the tech, we'll likely see a lot more cases like this one in the near future.
More on misinformation: That Viral Image of the Swagged out Pope Is an AI Fake, Dummies