On February 28, 2026, the United States and Israel struck Iran. Within hours, images began spreading across social media: destroyed cities, mass casualties, burning infrastructure, statements from officials.
Researchers identified an unprecedented volume of them as fake. AI-generated. Fabricated in seconds, distributed by coordinated accounts, amplified by algorithms that reward engagement regardless of truth.
Roughly two dozen X accounts — many with blue verification checkmarks that users associate with credibility — collectively accumulated over one billion views posting AI-generated content about the conflict. One billion. From two dozen accounts.
None of the content was real.
What a Billion Views of Disinformation Does
A billion views does not mean a billion people believed what they saw. But it means a billion encounters with fabricated content. A billion moments where someone’s understanding of a real, deadly geopolitical event was shaped by something that never happened.
The psychological research on misinformation exposure is unambiguous: even when people learn that content is fake after seeing it, the initial impression leaves a trace. It influences subsequent judgments. It creates doubt about true information that contradicts the false impression. The debunking never fully catches up with the original lie.
Now multiply that across a billion encounters. Across a conflict that could escalate into a wider regional war. Across a public that is already deeply polarized and deeply distrustful of official information.
This is what information warfare looks like in 2026. And there are no rules governing it.
The Technology Is Essentially Free Now
Generating a convincing AI image of a destroyed building — one that looks like it could be a real photograph from a conflict zone — takes approximately 30 seconds and costs fractions of a cent using current tools. Generating a realistic video of an official making a statement they never made takes minutes.
The accounts running these disinformation operations during the Iran conflict were not nation-states with intelligence budgets. The barrier to entry for large-scale AI disinformation is now accessible to individuals, small groups, political operatives, and anyone with a motivated agenda and a laptop.
The World Economic Forum identified AI-enabled disinformation as the single greatest global risk in its 2025 and 2026 risk assessments — above climate change, above economic instability, above military conflict. This is not alarmism. This is the assessment of the most comprehensive annual survey of global risk experts conducted anywhere in the world.
The Platforms Are Losing. Badly.
X, Meta, YouTube, and TikTok all have policies against AI-generated disinformation. They are all enforcing these policies inconsistently, inadequately, and always at least one step behind the latest generation of generation tools.
Elon Musk’s decision to reinstate suspended accounts, reduce content moderation staff by 80%, and sell blue checkmarks to anyone willing to pay $8 per month created precisely the infrastructure that disinformation operations needed to scale. Verified-looking accounts with no accountability, operating on a platform with gutted enforcement, in a global media environment where speed matters more than accuracy.
The result: 1 billion views of fake war content in one conflict, in one month, from two dozen accounts.
The International Fact-Checking Day Irony
April 2, 2026 was International Fact-Checking Day. News organizations published guides on how to spot AI-generated images. Educators created lesson plans. Researchers shared tools for detecting synthetic media.
Meanwhile, the AI systems generating disinformation improved faster than the detection tools trained to catch them. The gap between generation capability and detection capability is widening, not narrowing. The cat and mouse game has a fundamental asymmetry: generating convincing fake content is getting cheaper and easier, while detecting it requires expensive, specialized analysis that no platform is applying at scale.
What This Means for Every Conflict From Here Forward
The Iran situation is not an exception. It is a preview.
Every future conflict — military, political, economic — will take place simultaneously in physical reality and in an information environment where fabricated content is indistinguishable from real content to most viewers. Where the first images of any event may be AI-generated by actors with an agenda before any journalist can reach the scene. Where the emotional narrative of what happened is set by disinformation before the facts can be established.
This is not a problem that will be solved by media literacy campaigns or platform policy updates. The technology has outpaced the institutional responses. The rules that govern physical warfare — the laws of armed conflict, the protections for journalists, the prohibitions on targeting civilians — have no equivalent in information warfare.
One billion views of fake content about one conflict in one month. The next conflict will be bigger. The tools will be better. The rules will still not exist.