Since in retaliation, the battlefield has not just stayed on the ground but expanded to include the internet. A huge amount of disinformation is being spread there by both sides: mainly old videos being reused as allegedly current material. DW Fact check also found AI-generated videos created to spread false information.
debunked some of them and tells you what to watch out for and how to fact-check content yourself.
Claim: A video clip compilation shows multiple bombings from the air, explosions and fire at night. In the one-minute video shared on TikTok “No mercy – Israel launches revenge attack on Iran” is written in capital letters. In the caption, the user repeats the claim: “Israel bombing Iran with no mercy.” The video has been seen more than 2.8 million times.
DW Fact check: False
The videos do not show the current conflict between Iran and Israel. They are 22 years old and date back to the US bombings in Iraq’s capital, Baghdad, in 2003.
A reverse image search of stills from the compilation reveals two sources that prove the time and place for sequences of the video. The first sequence of the TikTok video is part of a longer video compilation of a bombing at night in Iraq, published by the US media outlet CNN (minute 2:46.)
A sequence that starts at minute 0:31 on TikTok can be found here at the Getty Images photo agency.
In conflict situations, old footage or pictures are often shared out of context. In this case, many users pointed out in the comment section that the videos show the US strikes in Baghdad. When in doubt whether a video shows a current conflict, the first clue can be to check other users’ comments. However, sometimes users misuse this to sow doubt and claim that authentic material is AI-generated or mislabeled.
This video is also an example of how unreliable AI chatbots are as a fact-checking tool, as DW Fact check . In the comment section under an X post containing this video, some users turned to Grok, X’s AI chatbot, in an attempt to verify what the video shows.
Grok incorrectly answered, it “likely shows Iranian missile strikes on Tel Aviv on June 16, 2025.” In one comment, Grok even contended that its answer is “based on reports from CNN, BBC, and The Guardian. These sources confirm similar events, with no evidence suggesting the video is from another location.” This is not true.
Even upon further questioning and providing information that the video shows Baghdad in 2003, the Grok AI chatbot sticks to its own theory but adds: “But claims it’s 2003 Iraq War footage can’t be fully dismissed—similar visuals exist, and mislabeling has happened before. Without clear proof, I’d say it’s probably recent but urge caution with social media clips.”
As DW Fact check has shown, there is clear proof that the videos date back to 2003.
The same or similar versions of the video have also been spread to claim it allegedly shows Iranian bombings of Israel. These claims appear here in Arabic or Urdu or also on the French version of Pravda, a website that is known to be part of a Russian disinformation campaign.
AI-generated content getting more and more realistic
Claim: A 16-second video clip shows a bird’s-eye view of a city with massive destruction and collapsed buildings. A location stamp in the video reads in Arabic, “Tel Aviv.” This video has been shared widely, for example, on X by the Iranian media outlet Tehran Times, calling it “Doomsday in Tel Aviv.” It has been viewed more than 1.9 million times. The video has been shared in Arabic as well, with the claim, “Unusual. This isn’t Gaza, this is Tel Aviv.”
DW Fact check: Fake
The video is not real; it is AI-generated. It doesn’t show destruction in Tel Aviv.
With a reverse image search, we found out that the video was published on TikTok on May 28, before the current escalation between Israel and Iran. The account’s bio reads “The AI resistance.” Many of the videos published on this account are AI-generated.
Two clear clues reveal that the video is AI-generated, which are marked in the screenshot below. In the back, two cars merge while driving towards each other at 0:07 of the video. In the lower part, the shades on top of the roofs give it away: The shadows of the cylindrical objects would need to be parallel as a rule of physics, as the sun, as the source of light, is so far away. However, they are not. On the left roof, the shadows point slightly more to the left than on the right roof.
In the top left corner, two cars merge while driving towards each other. At the bottom, the shadows of the cylindrical objects on top of the roof would need to be aligned at the same angle. On the left building, they turn more towards the left than on the right building.
According to AI expert Hany Farid, this video has been created by the AI model Veo3, as he stated in a post on the social network LinkedIn. Farid is a professor at the California UC Berkeley School of Information who works in the field of digital forensics. He is the co-founder of GetReal, a company that aims to detect AI deepfakes.
How to spot AI-generated videos?
Farid said AI models currently have limits to a maximum shot length of eight to 10 seconds for one video sequence. If a video has this maximum length or consists of multiple short shots, it is an indication that the video might be AI-generated.
It is becoming more and more difficult to judge with the naked eye whether a video is AI-generated. Tools have become better and better at creating very realistic videos. When in doubt, try to find a version of the video in higher resolution. Be cautious if a video looks grainy or has a low resolution. People who spread disinformation use downscaling to hide manipulation, as visual inconsistencies are much more difficult to spot in a grainy video.
Poor video quality can also indicate that a clip was downloaded and uploaded multiple times. This usually results in a lower video quality. In this case, the original video posted on TikTok has a much better resolution than the other versions.
Footage of chemical explosions in China misused
Claim: Many viral tweets on X, including this one, which received 2.1 million views at the time of writing, claim in Arabic that this video shows an Iranian bomb destroying arms stocks in Tel Aviv. It reads, “A huge Iranian bomb fell, destroying weapons depots in Tel Aviv, causing the largest explosion in the history of the occupying entity.”
Another post shared the same video and also claims it shows an Iranian strike in Israel. Here, the post claims it shows an explosion in the port of Haifa after Iranian strikes: “The blast at Haifa Port is set to be remembered in history as one of Iran’s most lethal and strategically devastating strikes. Witnesses were left in shock as the sheer scale of the explosion unfolded before their eyes.”
DW Fact check: False
The video doesn’t show an explosion in Israel caused by an Iranian attack. It does not show Haifa or Tel Aviv. A reverse image search reveals that the footage is much older and from a totally different region. It shows chemical explosions at a factory in the port of Tianjin in northern China. This incident occurred on August 12, 2015, killing 173 people according to Chinese authorities.
It is not the first time that this video has been presented in the wrong context. In 2023, the same video was falsely presented as a Turkish nuclear power plant on fire after the earthquake in February in Turkey and Syria.
Emad Hassan and Kathrin Wesolowski contributed to this report.
Edited by Rachel Baig
The post Fact check: Old war videos recycled as fake news appeared first on Deutsche Welle.
