Real or AI? The red flags to spot fake war footage online
Debunking false AI-generated Iran War content
When it comes to war, many of us have seen images of missile attacks, burning buildings, and attacks on U.S. bases that look suspicious and wondered if they're real. AI experts are helping decipher some of the images and videos that are depicting the war. FOX 13's Kylie Jones reports.
TAMPA, Fla. - Experts are helping debunk fake images and videos of the war with Iran that are circulating online.
Operation Epic Fury has been met with a counter-attack in the form of pro-Iranian AI-generated war images.
What we know:
AI-generated images and videos posted on social media platforms by Iranian state media and leadership were debunked on Fox News, after they depicted false narratives out of the Middle East.
For example, a video posted on "X" by Iranian state media falsely claims the largest U.S. radar in the Persian Gulf was the target of a missile strike.
A video posted on Facebook showed a skyscraper burning as civilians watched in what was said to be Bahrain, but at the bottom of the video, you could see morphing cars.
One video falsely depicted missiles raining down on Tel Aviv.
A fiery image was posted on "X" that incorrectly claimed a U.S. base was up in flames in Iraq.
What they're saying:
U.S. Central Command has taken to social media to fact-check many of these posts and other claims by Iranian leadership.
U.S. officials say many of these posts and statements are lies, misleading warnings and false narratives.
Dig deeper:
"It becomes easier and easier to create AI-generated content," Dr. Jill Schiefelbein, an AI strategist said.
Schiefelbein says there are a number of ways you can quickly determine whether certain content is real as you scroll.
READ: Florida Poly to launch cybersecurity and AI laboratory on campus
"Note the source," she said. "Is this actually a reputable, trustable source? Is this something that you could actually go and verify?"
Other AI experts say there are certain characteristics that are common giveaways with AI-generated images and videos.
"If it's too perfect, like it's a clear explosion, or sometimes you look at the little details, like sometimes you might look at uniforms, and the spelling is a little bit wrong, because AI always messes up with the wording at times," Mohamed Yousuf, the CEO of Smart Workforce AI said. "You can see that if the lighting is too perfect, especially in a war zone, or if the camera's held exactly at the perfect angle."
Big picture view:
Yousuf says you can also pay attention to whether the account posting the content is a new account.
"The problem is, the algorithm doesn't help either, because a lot of people are viewing something, it's going to increase the view count, increase the likes, and then you have people sharing fake realities," he said.
READ: Florida Legislature moving in different directions on AI regulations
AI-generated content can be created in seconds.
"Pay attention to their body language, as well," Schiefelbein said. "What you'll find in a lot of AI-generated videos is that you have very homogenized backgrounds, very similar acting agents in the background. And then the foreground changes a bit."
What you can do:
Schiefelbein says there are also systems that you can run images and videos through to detect if they were generated by AI.
"You can also use different types of tools that will examine the veracity of these images," she said.
Experts say it's important that users help report fake content when they come across it.
They also say social platforms should have better regulations and controls in place to help prevent fake content from being uploaded.
"When we can get a system where creators and account holders are penalized for uploading AI-generated content without disclosing that it's AI-generated content, we will start coming to a better solution," Schiefelbein said.
The Source: Information for this story came from interviews with AI experts and information shared by U.S. Central Command.