In a predictable but dark example that sheds light on the future of the internet, a fake photo of the Pentagon on fire caused the US stock market to plummet briefly, once again demonstrating how important it is to contain the artificial intelligence-driven misinformation crisis.
The image shared by a verified account called “Bloomberg Feed” (that is, paying $8 for the blue tick) showed smoke rising from the Pentagon after an alleged explosion, prompting widespread speculation about a possible attack. The headline read, “The Big Bang near the Pentagon Complex in Washington DC – First Report” and was clearly trying to imitate a genuine Bloomberg article. However, as you can imagine, the account was completely fake and this incident once again highlights the ongoing problem with Twitter’s paid verification process.
The tweet, which was shared by famous Twitter users with a high number of followers shortly after, caused the image to spread rapidly on social media and caused panic. It was even spread by RT, a Russian state media account with 3 million followers, and caught the attention of the markets.
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
DeItaone, a Twitter user with 650,000 followers, shared the tweet at 10:06 a.m. local time, and stocks fell 0.26 percent at 10:10 a.m. just four minutes later, Insider reported. Although this number is not huge, its realization in such a short period of time indicates that it may have effectively frightened the markets.
Then the markets rebounded and most of the tweets sharing the content were deleted, including both RT’s and DeItaone’s. However, this event provides an effective opportunity to examine a situation that may be even more worrying in the future.
The image had all the hallmarks of an AI-generated image, the building’s columns were in the wrong places, the fence was strangely disappearing into space, and the windows were not aligned. So it seems likely that someone connected a prompt to a productive AI and shared it as a fake story. This event is just one of the things experts have been worried about since ChatGPT, Midjourney and more demonstrated their extensive capabilities.
The incident doesn’t just point to the alarming capabilities of artificial intelligence, either. It also illustrates the ramifications of the Twitter verification system becoming a subscription service with a simple payment of just $8.
56551
