In a predictable turn of events, the combination of paid blue checks and generative AI has made it alarmingly easy to propagate misinformation. On Monday morning, an AI-generated image depicting an explosion at the Pentagon began circulating on the internet, despite the fact that no such event had occurred.
Within a short span of approximately 30 minutes, the image surfaced on a verified Twitter account named “Bloomberg Feed,” which could easily be mistaken for an authentic Bloomberg-affiliated account due to its blue checkmark. However, the account has since been suspended. Screenshots captured by users reveal that the Russian state-controlled news network RT also shared the image before deleting the tweet. Additionally, several Twitter accounts with substantial followings, such as DeItaone, OSINTdefender, and Whale Chart, spread the image. Even an Indian television network reported the fabricated Pentagon explosion. The origin of this fake image and news story remains unclear.
While the circulation of fake images is not a new phenomenon, the stakes are higher when the false event involves an explosion at a U.S. government building rather than a trivial matter. There have been suggestions that the fake image may have influenced a slight 25 basis point movement in the S&P 500, but the dip was short-lived, and it cannot be definitively proven that it resulted solely from this hoax. However, this incident does raise concerns about the potential future use of generative AI to manipulate the stock market, as previously witnessed with Reddit.
Misinformation has long been a challenge on the internet, but the simultaneous growth of generative AI and changes in Twitter’s verification system have created fertile ground for its spread. Twitter owner Elon Musk’s plan to remove blue checkmarks from existing accounts and allow anyone to pay for verification has proven to be a chaotic endeavor from the start.
Although it is known that blue checkmarks no longer guarantee legitimacy, breaking the visual habit cultivated over nearly 15 years is challenging. If confronted with an account named “Bloomberg Feed” that boasts a blue checkmark and posts about an attack on the Pentagon, one is predisposed to assume its authenticity. As the identification of fake images becomes increasingly difficult, false news reports of this nature will persist in the future.
