When catastrophic flash flooding occurred in Nepal in late August, videos of the disaster almost immediately began surfacing online. Videos showed biblical floodwaters tossing charter buses like toys, destroying concrete buildings and mighty bridges and quickly overwhelming people desperately trying to get away.
While some of these videos were authentic depictions of what was happening in Nepal, others shared online weren’t. They were either generated with artificial intelligence — or not from the Nepal flood at all.
Download the Straight Arrow app today to get the stories that matter free from manipulation, bias or agenda.™
Point phone camera here
The proliferation of these videos raises questions about what people get from sharing fake or misleading content in the wake of a major disaster.
One of these inauthentic videos posted on X, reviewed by the AFP, shows a large rush of floodwaters breaching a riverbank before spilling onto a road, submerging cars and people. But a closer look shows several irregularities, like a man vanishing from the video just as the waters spill onto the street. X later added a community note saying the video was created using AI.
Other users posted non-AI videos but falsely claimed they were from the Nepal floods, according to France 24’s Truth of Fake segment. In one of these videos, users claimed a Chinese helicopter filmed a glacier where the floodwaters that swarmed Nepal originated. The video was actually filmed in Alaska, more than 5,000 miles away.
What people gain from creating fake content
The motivation for creating fake content, especially after a disaster, is simple: money.
Social media platforms often pay creators based on how many views a video receives, regardless of whether it’s authentic, according to 404 Media.
During the Los Angeles fires in early 2025, content creators quickly began rolling out AI-generated videos appearing to show famous monuments on fire and firefighters saving cute animals from the inferno, 404 Media said. None of them were real, but the videos quickly racked up millions of views.
Meta, which owns Instagram, Facebook and Threads, hasn’t publicly said how much it pays creators, but 404 Media quoted influencers who said they receive $100 to $120 per million views. Earnings from a viral video can quickly add up.
One content creator the publication cited posted several AI-generated videos in a 24-hour period. Within a day, the creator got about 94 million views — worth about $9,400, using 404 Media’s lower revenue estimates.
Getting millions of eyes on content can also lead to more followers for creators, which could lead to more consistent viewers and still more money. Some creators, like the one 404 Media highlighted, even create paid courses to teach others how to produce AI content that makes money.
Why disasters are ripe for AI content
Many content creators publish AI videos after disasters because viewers are more likely to share material that plays on emotions. According to a study published in PNAS Nexus in 2023, researchers found that posts on X that used negative emotional language were more often shared than posts with less negative language.
Another study published in Scientific Reports a year later found similar results. Researchers found that users on X and Facebook were 1.91 times more likely to share a negative news article compared to less negative content.
Most social media platforms typically add an AI content disclaimer to posts they believe were generated with the technology. Occasionally they go further and add a community note disclosing the same. But those warnings can be easy to miss.
TikTok has the most explicit rule regarding disasters, banning any synthetic media depicting “fake authoritative sources or crisis events,” according to its website. Still, fake disaster content is shared prolifically on social platforms.
This is because when an event like a flood, wildfire or earthquake happens, people want answers fast. But real, verified reports are slower to reach the public, leaving people more vulnerable to a quickly generated video or a real video of a past event, according to Digital Trends. The outlet described it as “an ugly race” in which viral videos appear before reliable context.
Sharing is caring
AI-generated videos and misinformation are not only shared by the people who create them but also by people who discover the content. Psychologists have a few explanations for why people share things that aren’t true.
One explanation is about belief, according to a summary by the American Psychological Association. The APA said more exposure to misinformation increases the odds that someone will believe it and that belief increases the odds they’ll share it.
But belief isn’t the full picture because sometimes people share things they know are false. People might do that to show their political side or put the other side down, the APA said. Other times it’s just to look good in front of their digital friends. The association also said people are more likely to share something if it fits who they are, feels new or stirs up strong emotions.
Other research tells a different story. A group of psychologists led by Gordon Pennycook and David Rand ran a series of experiments on this and published their findings in Nature in 2021. They found people are often good at spotting false headlines when asked directly. But that doesn’t carry over to what people actually share.
Even people who said accuracy mattered most to them still said they’d share 27.7% of the false headlines they saw. The researchers don’t think belief is the real problem, but attention is. Rather than thinking about whether something is true, people focus on other factors, like how it makes them look.
Applying this to disasters like the Nepal floods, both theories point to the same result, even if they disagree on the cause. A disaster is emotional and feels urgent. Fake depictions of the event also move faster than real, verified reports. No matter the reason, dramatic footage continues to spread faster than anyone can confirm and socialize its authenticity.
Round out your reading
- Former Fetterman staffers create social media account to expose former boss.
- Why Gen Z finally wants Boomer hand-me-downs.
- Want a copy of Pete Hegseth’s driver’s license? It’s for sale on the dark web.
- Smart glasses helped a man see the world. The privacy panic took it away.
- Americans are keeping cars longer, so the auto industry is making it easier to pay for repairs.
