Imagine living in a world where you can’t trust your own eyes. Where a devastating bombing in a distant city, shared widely online, could be a complete fabrication. This isn’t some far-off dystopian novel; it’s our reality right now, thanks to the explosion of Artificial Intelligence (AI) and something called “deepfakes.” Just recently, a video supposedly showing missiles hitting Tel Aviv in March 2026 went viral. It looked incredibly real, but it wasn’t. It was an AI-generated fake, designed to look like a terrible event that never actually happened. This kind of digital trickery is now everywhere, especially with recent conflicts heating up, like the rekindled military actions between the U.S. and Israel with Iran. Our social media feeds are drowning in these AI-created fakes: videos of fake celebrations, panicked airport scenes, and horrific casualties. It’s like a constant firehose of misinformation, making it nearly impossible to tell what’s real and what’s make-believe. This isn’t just about entertainment; it’s about understanding the world around us. Because of this, learning to be “Critically AI Literate” (CAIL) isn’t a bonus skill anymore; it’s absolutely essential. We’re no longer in a “fog of war” that’s just confusing; we’re in an information environment literally choked by “AI slop.” Shockingly, over 20% of YouTube content is AI-generated, and without a serious effort to teach people how to spot these fakes and understand the forces behind them, we’re completely exposed to sophisticated mind games. We need to go beyond simply knowing how to use AI; we need to understand the powerful forces that own this technology, and the hidden biases they embed within it.
The idea of using false information to manipulate people isn’t new; it’s as old as conflict itself. Think about the Trojan Horse, a massive wooden horse the Greeks used to sneak into Troy and win a war. Or how Genghis Khan’s warriors would pretend to retreat, drawing their enemies into a deadly trap. Deception has always been a key weapon on the battlefield. In modern times, especially in democracies, leaders have tweaked these old tricks into “false news” to get public approval for wars. We saw this with the so-called “phantom” attack in the Gulf of Tonkin, which was used to escalate the Vietnam War. And who can forget the “phantom” Weapons of Mass Destruction (WMDs) that were trotted out to justify the 2003 invasion of Iraq? But misinformation isn’t just for starting wars; it’s also used to keep spirits high and pretend things are going well. During the Vietnam War, the White House kept telling everyone the U.S. was winning, even as internal reports painted a grim picture of a worsening situation. Similarly, President George W. Bush declared “Mission Accomplished” just weeks into the Iraq War from an aircraft carrier, giving a false sense of triumph to a conflict that would drag on for decades. It’s a recurring pattern: spin a false narrative, get people on board, and then keep them there with more false hope, even if the reality is far more complicated and tragic.
What’s different now is the sheer power of AI and social media. While the desire to deceive is ancient, these technologies have thrown gasoline on the fire, allowing anyone to crank out incredibly believable, polished fake content at an unprecedented scale. Even before the recent conflicts flared up, the wars in Ukraine and the tensions between Israel and Bahrain were already swamped with AI-generated misinformation. The spread of these “deepfakes” does more than just spread lies; it’s eating away at our basic ability to believe in objective truth. When everything can be faked, everything becomes suspect. This creates a dangerous situation where even genuine evidence of suffering can be dismissed as just another trick. For example, NBC News had to conduct a painstaking investigation to confirm that a video of starving Gazans waiting for food in May 2025 was real. Despite the journalistic rigor, a flood of social media users instantly dismissed it as a deepfake. When people can no longer tell the difference between a clever fake and documented reality, truth becomes less about facts and more about what serves someone’s political agenda. In emotionally charged situations, this “fog of war” can whip people into a frenzy, making them feel like their lives depend on quick decisions. If they’re acting on bad information, a peaceful protest could turn into violent extremism. Social media platforms, in their quest for engagement, actually reward this chaos. Fake news, being more sensational and provocative, often spreads faster and wider than the complex, nuanced truth.
It’s tempting to think that people should just pause and investigate before believing everything they see, especially when it looks like a massacre is happening. But that’s a huge ask. While some deepfakes are easy to spot (like a video of Israeli Prime Minister Benjamin Netanyahu with six fingers – a genuine error in an AI generation), verifying information often takes time and expertise. You need to geolocate footage, check digital fingerprints (metadata), and sometimes, you just have to accept that there isn’t enough evidence yet to be certain. AI has made this truth-seeking mission incredibly difficult for the average person who doesn’t have the tools or know-how for deep digital forensics. Ironically, many people are now turning to AI itself to tell them if something is AI-generated, which shows a deep lack of understanding about what AI actually is. What we call “AI” today are mostly Large Language Models (LLMs). These aren’t genuinely “intelligent” in the way humans are; they’re more like incredibly sophisticated pattern-matchers that predict the next best word or image based on vast amounts of data. They’re only as good as the data they’re trained on and often amplify human biases to a dangerous degree. Studies consistently show that AI responses can be flat-out wrong about half the time, often “hallucinating” facts and sources that don’t exist. The Intercept highlighted this absurdity by showing how Google’s Gemini AI gave contradictory answers about whether a piece of text was AI-generated, even when Gemini itself had created that very text! When news organizations use AI detectors as definitive proof, they’re effectively building their conclusions on quicksand.
This widespread AI illiteracy is built on decades of neglected media literacy. While many countries have made media literacy a required part of their education, the U.S. has largely left it up to individual communities. Media literacy is about understanding how to find, analyze, evaluate, create, and use all kinds of communication, from newspapers to complex digital media. Without this basic foundation, people are completely unprepared for the complexities of an algorithm-driven world. “Critical AI Literacy” (CAIL) goes beyond just knowing how to use a chatbot. It teaches you to ask fundamental questions about power: Who owns this AI? How does that ownership influence its biases, its underlying beliefs, and its ultimate purpose? If a giant corporation owns a powerful AI model, will it prioritize making money over stable, democratic societies? CAIL also makes us examine representation. It prompts us to consider how AI-generated images reflect the biases embedded in their training data – sometimes even surfacing white supremacist or extremist content from unmoderated models like Grok AI. It reminds us that often, the tech industry, in its core philosophy, can be anti-human, viewing people as flawed systems that need to be “fixed” or “optimized” by code, rather than valuing our inherent complexities.
As researcher Gary Smith wisely put it, AI will only surpass human intelligence if we, as humans, continue to use it in ways that degrade our own cognitive abilities. Evidence shows that constantly relying on AI and screens without thinking critically can dull our minds, hurt our memories, and shorten our attention spans. CAIL’s core message is empowering: humans are the smart ones; AI platforms are just tools. In times of war, the absence of this literacy can have deadly consequences. If deepfakes and AI “hallucinations” are manipulating our emotions and shaping how we understand global conflicts, we are trapped in a never-ending, synthetic crisis. We simply cannot afford to repeat the mistakes of the past, where we naively believed that simply having access to technology would automatically make the world more connected and smarter. The whole point of Critical AI Literacy isn’t to make us fear or reject technology. It’s about empowering us to understand it deeply, so we can use it for the benefit of everyone. We have a fundamental choice to make: Will AI be a partner that streamlines boring tasks and genuinely improves human lives, or will it become an exploitative force that dictates what citizens perceive as reality? This crucial decision belongs to an informed public, not powerful tech executives. If people remain ignorant about AI, they will remain trapped by the very narratives designed to exploit and control them.