AI-Generated Dam Collapse Video Spreads After Nepal-Tibet Flood

A fabricated video purporting to show a catastrophic dam failure near the Nepal-Tibet border spread rapidly on social media following deadly flooding in late August 2026, demonstrating how AI-generated content can distort public understanding of natural disasters.

The video, shared widely on Facebook, Instagram, and X, claimed to depict the collapse of a Chinese-built dam near Gyirong Port that "triggered flash floods and endangered the lives of hundreds of thousands of people." The footage appeared to show a massive dam breaking apart and releasing torrents of water toward populated areas.

But the video was fabricated, and the disaster's actual cause was entirely different, according to AFP's fact-checking investigation.

The real disaster

The flooding that struck the Nepal-Tibet border region on August 26, 2026, killed at least 1,385 people and left more than 5,500 missing. However, the US Geological Survey determined the catastrophe resulted from a high-altitude glacial collapse that sent ice and rock debris surging downstream—not from any dam failure.

Nepal's foreign minister addressed the true cause on August 29, warning about mounting climate threats in the Himalayas. UN climate chief Simon Stiell echoed this concern, calling the event "a devastating reminder of how fragile these mountain environments can be."

Telltale signs of AI generation

AFP's analysis revealed multiple visual anomalies characteristic of AI-generated media throughout the circulating footage. In the opening sequence, a man running toward the supposed dam appears to morph into a different person. The dam itself, despite allegedly collapsing, remains largely intact in the video.

In footage purportedly showing flooding from downstream, a black car vanishes at the 24-second mark. People standing on rooftops display unnatural movements and show no reaction to the chaos supposedly unfolding around them.

Meta's AI detection tool confirmed the video contained digital signatures indicating it was likely created using artificial intelligence tools.

Why it matters

The incident illustrates a dangerous evolution in disaster misinformation. AI-generated videos can now create convincing false narratives about catastrophic events in real time, potentially misdirecting relief efforts, inflaming geopolitical tensions, and obscuring the actual causes of disasters—in this case, climate-driven glacial instability. As extreme weather events increase, the ability to quickly distinguish authentic documentation from fabricated content becomes critical for emergency response and public safety.

The details in this report were first published by AFP's fact-checking team.