Photo of a massive explosion at an airport in Iraq. Satellite image showing damage to a US naval base in Qatar. Video of an Iranian ballistic missile hitting central Tel Aviv. These are all images that have been circulating in the past week since the Trump administration attacked Iran. And none of them are real.
These images, along with many others, were created or manipulated by AI to spread misinformation about what is actually happening in and around Iran, and are becoming increasingly problematic for those seeking to separate truth and reality from lies and propaganda.
Spreading misinformation has always been a part of war, as opposing sides vie for public support while launching bombs. But now, with generative AI, the ability to fake images and videos is easier than ever. Gone are the days when you needed Photoshop skills to create a false narrative. And with social media, these doctored images can be sent across countries in seconds. While malicious actors may be intentionally trying to sow discord, the number of people who unknowingly share discord is increasing exponentially. This, combined with the White House’s intent to spread propaganda, creates an information ecosystem that feels overwhelming and confusing.
“Deepfakes of video, audio, and images have reached a level of realism that is indistinguishable from reality for most people,” says Raman Chaudhry, a prominent AI researcher and former head of ethics at X (when it was still known as Twitter). “AI companies are agreeing to watermarks and other verification methods, but they are not built with how users interact with social media.”
“This is especially dangerous given a situation like the Iran war,” Chaudhry added. “Most Americans are likely entering the country with information that is ill-informed and likely biased. Fake media only confuses and exacerbates these biases.”
“Shallow Fake”
February 28, Iranian state newspaper tehran times Shared satellite images purportedly showing destruction left behind after an Iranian drone struck a U.S. military radar installation at a U.S. naval base in Qatar.
BBC Verify, a team of journalists specializing in image fact-checking that tracks and maps attacks related to the US-Israel war on Iran, labeled this satellite image as an AI-generated fake. The research team explained that the altered images were based on actual satellite images of US military bases, but were edited with Google AI to falsely depict the damage.
These days, it’s not as simple as people being fed or fooled by deepfakes. Political scientist Steven Feldstein said that as people became more knowledgeable about AI, the creators of disinformation content also became more sophisticated in the way they presented things, resulting in more sophisticated manipulations, or “shallow fakes.”
“Instead of presenting something that looks like a complete lie, [they] “It’s about presenting shades of truth and manipulating what’s there,” Feldstein says. That is, content creators provide enough detail and nuance to slip past people’s bullshit detectors, but can still misinterpret reality and represent a particular point of view. This can happen with just a little manipulation of the image. For example, a photo of a real Iraqi airport depicting smoke above a US military base in Iraq was altered using AI to show a giant fireball explosion on March 1st. Or you may be sharing out-of-context images such as: I used an old photo and said it just happened.
“You see it happen as the levels go up,” says author Feldstein of The rise of digital oppression: How technology is reshaping power, politics, and resistance. “It’s very sophisticated and it’s also an important part of geopolitics.”
Feldstein said the 12-day war in June 2025, when Israel and the US attacked Iran, showed how quickly AI-generated content can spread. BBC Verify’s Shayan Sardarizadeh told the Global Investigative Journalism Network that the attack was “the first example of a major global conflict where we have seen more misinformation being generated using AI than traditional methods”. Sardarizadeh said this “ushers in a new era in how AI-generated content is used in the wake of major breaking news,” adding that his team has confirmed that some AI-generated videos and images have been viewed “millions of times” and mislead people.
Chaudhry, a former US science envoy for AI, said there are already “disinformation agents on social media promoting specific agendas.” She pointed to when X released a location feature in November of this year that uses IP addresses to show where an account is based. “We found that many of America’s right-wing influencers are located in Africa, Bangladesh, Russia, and Ukraine.”
Both Chaudhry and Feldstein say the blurring of fiction and reality can lead people to claim it’s fake when they see a real video. When you can no longer trust your own eyes, it becomes difficult to challenge your strongly held beliefs.
“We’ve now reached the point where we can’t accept anything that comes in beyond our own existing narrative as truth, and that’s just as harmful,” says Feldstein.
“War is not a video game”
Misinformation doesn’t always rely solely on AI. For example, video game screenshots are sometimes circulated and shared as if they were actual photos of destruction. And then there’s overt propaganda, which has grown exponentially under the Trump administration. During the ICE raid in Minnesota, the Trump administration relied on cruel memes and AI-generated images to try to sway public opinion. They also created their own version of shallow fakes, digitally altering photos of arrested women to make it look like they were crying when they weren’t.
Then came Iran. On March 4, the White House released a video on its official X account that combined actual clips of Iranian missile attacks with Iranian footage. call of duty video game. Halfway through, a choppy voiceover says, “We’re winning this battle.”
On March 5, the White House released another video, this time praising “American justice” with clips from movies and TV series, including: brave heart, breaking bad, iron man, and gladiator.
According to Iranian state media, more than 1,000 people have already been killed in the war, including more than 100 Iranian schoolgirls and at least six American soldiers.
“War is not a video game,” tweeted veteran and Barstool Sports podcaster Connor Crehan. “The outcome of war is final. I wish we had not treated war with such cavalierness.”
Feldstein said information and images are increasingly being used to call for action, and thanks to social media, rhetoric spreads so quickly that no one can stop and identify where the content is coming from. Without knowing who is making these claims and whether they are reliable news sources, it is difficult to determine whether the story being claimed is one-sided and from a particular point of view that may be challenged. He added that the US president and Israeli prime minister have used images and motifs to call on Iranians to rise up against the government. “That’s not the case in the United States.” [currently putting] “When it comes to the Iranian government, it relies on communication as a way to effect change on the ground, and you can see how high the stakes are when it comes to how quickly that information is digested and acted upon,” Feldstein said. [spurring] action. ”
And of course, he points out, there are huge humanitarian risks to bad information that go beyond political manipulation. People living in areas of armed conflict need to know where they can safely evacuate and where drones are attacking them if they need to evacuate. It is important to have reliable information.
