Impressive images showing the result of an explosion near the Pentagon circulated widely on Twitter yesterday.
Still, the United States Army Command Center is intact and in great shape, but unfortunately the same cannot be said for Twitter’s moderation services. Indeed, the image that quickly went viral was actually generated by artificial intelligence. Of the many accounts that ran it, many got their credibility from a certain $8 a month blue badge.
Where does this image come from?
The morning of May 22 was bound to be like many others for Pentagon officials, until a violent explosion near the building rocked them. At least that’s what anyone who wasn’t there could have believed when they saw the massively shared image from a Twitter account sporting Twitter Blue’s paid verification badge, also with a believable name: Bloomberg Feed. This “information” was quickly amplified by many other accounts with similar characteristics and counting hundreds of thousands of subscribers. Even an Indian television channel took over the information. Note also the sharing of the image by the media RTaffiliated with the Kremlin, which has since deleted the affected tweet.
It was neither Twitter nor even AI or misinformation experts who reacted first, but the emergency and fire services who ultimately denied the information. These were then retweeted by the Pentagon’s federal protection agency, itself devoid of the famous verification badge. Eventually, Twitter decided to take action, such as suspending the account Bloomberg Feed.
An idea of the future of disinformation
The consequences of this case will probably be minimal, but they are at the crossroads of two of the most important concerns among observers of disinformation. On the one hand, we find the use of artificial intelligence to create texts, videos or, in this case, images with the aim of misinforming. On the other hand, the paid verification system, Elon Musk’s flagship idea when he took over the social network, shows its flaws. The first launch of Twitter Blue had also been a day of celebration for usurpers, disinformants and many jokers on the social network, which had prompted the businessman to come to his senses to install guards. crazy.
Obviously, these are insufficient. If this case had a consequence, it is above all to show how easy it could be to circumvent them. Let us recall in passing that this repeated fiasco has nevertheless been emulated, Meta having launched in its wake similar offers for Facebook and Instagram. The future of information on Twitter is a little more twilight every day.
Sources: TechCrunch, Bleeping Computer
6