“Wow, I Visited There Last Vacation”… Dubai in Flames: Fake Inferno Created by AI Hallucination

by Kong Byeongseon

Published 14 Mar.2026 07:00(KST)

Updated 14 Mar.2026 09:12(KST)

The Burj Khalifa, the skyscraper in Dubai, United Arab Emirates, appeared engulfed in flames. Black smoke was seen rising into the sky from the building. In addition, a high-rise in Bahrain was also shown catching fire and beginning to collapse following an attack by Iran. People watched as the buildings burned. However, all of these scenes were actually videos created by AI. Upon learning that these videos were fake, people responded by saying, “It’s so easy for the public to be deceived by videos on social networking services (SNS),” and, “Our society is heading toward a dystopia-a bleak future.”

An image of a U.S. military base that spread on the social networking service X (formerly Twitter) earlier this month. The original image only showed black smoke, but it was manipulated with AI to look as if engulfed in flames. Photo by X capture

On March 11 (local time), major U.S. media outlets such as CNN and AP reported that photos and videos related to war, created with generative AI, were spreading rampantly on social networking services. The videos depicted famous locations in Middle Eastern countries set ablaze by Iranian attacks or people fleeing from bombs. However, AP explained that the videos exhibited signs of AI hallucination, such as buildings appearing to be duplicated or overlapped.

In reality, it is easy to find AI-generated content on SNS. A simple search for Dubai on TikTok brings up a video supposedly showing Dubai International Airport damaged by an Iranian drone strike. The video had around 35,000 views. The TikTok account tagged the video as “news” and stated, “This is the interior of Dubai International Airport after the damage,” adding, “The Dubai International Airport publicity office has launched an emergency response team in cooperation with authorities.” Despite the fact that this was an AI-generated video, people continued to debate in the comments over whether it was real or AI-made.

As AI technology advances, video manipulation is becoming increasingly sophisticated. Last month, China’s media company ByteDance unveiled its AI video generation model “ShiDance 2.0.” The company sparked controversy by using ShiDance 2.0 to create a scene of famous actors Brad Pitt and Tom Cruise fighting in an abandoned building. The AI agent platform “GenSpark” can also easily generate videos from text descriptions alone, with the creation process taking as little as 15 minutes.

The problem is that AI technology is being used to produce manipulated videos. Some create fabricated war-related videos to generate profit or to steer public opinion in a desired direction. Evangelos Kanoulas, a professor at the University of Amsterdam, explained, “Advanced AI technology can subtly alter textures, faces, lighting, or background details to create images that are even more realistic than the originals,” adding, “This serves to reinforce certain narratives about specific events.”

There are also warnings that one must be aware of the possibility of AI manipulation when consuming information on SNS. Hany Farid, a professor at the University of California, told CNN, “The guidelines for identifying AI manipulation from just a few months ago are already obsolete,” and emphasized, “Instead of just skimming posts on SNS, people should get their news from reliable outlets like the press.”