These AI-generated scenes from Dubai (left) and Tel Aviv (right) are both totally fake.
A US aircraft carrier destroyed by Iranian missiles. American bombs leveling a nuclear power plant. The Burj Khalifa engulfed in fire.
None of it happened, but that didn’t stop people from spreading fake videos online.
In the days since Trump’s weekend strikes on Iran, AI-generated videos realistically depicting entirely fabricated events have been spreading like wildfire on X and other social media platforms.
For years, X (formerly Twitter) was one of the most valuable tools for real-time information during breaking news events. But that era seems to be over. Since Elon Musk’s takeover of the company, the platform’s usefulness as a reliable news source has steadily eroded. Moderation has been gutted, the algorithm rewards engagement over accuracy, and resources funneled into their own, uniquely problematic AI platform: Grok.
In a disturbing sign of how deep the problem goes, Grok—X’s own AI tool—has been misidentifying AI-generated content as real.
We’ve seen political AI content creep into the platform before—like a fabricated video depicting Jake Paul at Iranian protests—but the recent strikes in Venezuela, and now Iran, have unleashed an onslaught of misleading AI video content.
The motivations behind the content vary. Some creators appear to be celebrating, using AI video to glorify Trump and Netanyahu’s military actions. Others seem aimed at manufacturing doubt about the war, undermining American public confidence, and muddying the information environment so badly that no one knows what’s real.
And in a disturbing sign of how deep the problem goes, Grok—X’s own built-in AI tool—has been misidentifying AI-generated content as real. (A spokesperson for X didn’t immediately address a request for comment, but shared links to recent posts by the firm’s safety team.)
Last month, Mother Jones’ Arianna Coghill spoke with AI content expert Jeremy Carrasco about exactly this kind of scenario. Carrasco finds the fake content concerning, but says the deeper harm is what this flood of AI content does to our relationship with real video. When fake footage is convincing and common enough, people start doubting everything—including authentic footage of things that actually happened. That’s the environment we’re now operating in.
Staying informed has never been more important, but in this moment, that means being particularly careful about what you accept as real—even if you think you see it with your own eyes.
