It started with a spark on a windy Hawaiian afternoon—a broken power pole, dry grass, and a fire that would go on to kill more than 100 people and erase a historic community from the map. But in the days after the 2023 Maui wildfires, another kind of fire was spreading, one that moved through phone screens and comment sections faster than any flame. Across Facebook, X, TikTok, and YouTube, thousands of people insisted that the fire wasn’t an accident at all. It was a “direct energy weapon,” they said, a laser aimed from somewhere in the sky, part of a plot to clear the land for a high-speed rail line or a futuristic “smart city.” Some said the government had used a “weather weapon.” Others claimed that FEMA was waiting to seize the homes of anyone who applied for emergency aid—a terrifying idea in a community with deep Indigenous roots and painful memories of colonization. These stories were false, but they were everywhere. Right-wing influencers amplified them. Accounts linked to foreign adversaries pushed their own versions. And in the middle of it all, exhausted emergency managers watched helplessly as their official information got drowned out by a flood of speculation, suspicion, and outright lies. The question at the heart of this chaos is simple, and it’s one that haunts every disaster response: who do you trust when your world is burning?
The problem isn’t just that misinformation exists—it’s that the conditions for it have never been better. Extreme weather events have always been chaotic, and rumors have always swirled in their wake. But climate change has made disasters more frequent, more ferocious, and more disorienting. At the same time, the information landscape has fractured into a thousand pieces. Major social media companies have abandoned fact-checking. Politicians distort and weaponize disasters for short-term gain. Government agencies are stretched thin. And generative AI has created a whole new way for people to find information that may or may not be true. A recent report from the News Literacy Project found that most teens encounter conspiracy theories on social media—and most of those teens are inclined to believe them. The result is what experts call a perfect storm. In an emergency, people need facts fast, but reliable data is often impossible to find. They turn to their feeds, where fear spreads faster than truth. And that has real consequences. A Department of Homeland Security study found that an inaccurate map tweeted during a 2017 California flood caused a needless mass evacuation. A study from the Inter-American Development Bank found that an earthquake in Mexico led to an 11 percent drop in trust in government. A review of recent research concluded that bad information often leads to wasted resources, overwhelmed emergency services, and a dangerous diversion of attention away from genuine needs. In other words, the chaos we see online doesn’t stay online. It slows relief. It erodes trust. It costs lives.
Nowhere is that mistrust more visible than in the way people view the government—especially FEMA. The Federal Emergency Management Agency has been a magnet for conspiracy theories since it was created. There are the infamous claims about “FEMA camps,” where survivors would allegedly be imprisoned by the government. There are the endless variations on the same theme: that federal aid is a trap, that officials will take your land, that the agency’s real mission is control, not help. Deanne Criswell, who led FEMA under President Biden, says this stigma has haunted the agency from the very beginning. “It’s this idea of the black arm of government,” she said, and it makes it harder to get help to people who need it because they don’t know who to trust. During Hurricane Helene in 2024, this distrust turned dangerous. A rumor spread that a man had arrived in North Carolina to shoot FEMA workers, forcing Criswell to shut down field operations for hours. Around the same time, she and her deputy chief of staff were “swatted”—anonymous trolls sent police to their homes with false accusations, triggering SWAT deployments. The election season made everything worse. Right-wing accounts falsely accused Biden of starving North Carolina of aid. Elon Musk claimed FEMA was blocking him from flying in to help, then retracted the claim after his initial post had already been viewed nearly 30 million times. But mistrust of government isn’t always political. Some people ignore evacuation orders because they’ve been through false alarms before. Some don’t seek help because they’re undocumented and afraid of being arrested. Some communities have generations of painful experience with law enforcement. As one expert put it, there are many reasons people can’t or won’t trust the authorities, and all of them are hard to overcome.
Social media is the soil in which all this misinformation grows. Rumors about disasters are nothing new—after the Johnstown flood in 1889, people accused wealthy industrialists of causing the deaths of thousands; after Hurricane Katrina, claims of concentration camps circulated. But platforms like Facebook, Instagram, and X have made these stories global in minutes. They’re almost impossible to trace, and they’re rewarded by algorithms that favor outrage. A study following a 2023 earthquake in Turkey found that about a quarter of residents believed some version of the conspiracy theory that the United States caused it with a seismic weapon—and that belief often lined up with support for the incumbent government. The Center for Countering Digital Hate found that most conspiracy posts came from verified accounts. Even harmless fakes, like the famous doctored image of a shark swimming on a flooded highway, fooled media outlets and even a U.S. senator. These stories don’t just make people confused; they change behavior. In Japan, one study found that communities with low trust in their government evacuated even when they didn’t need to, clogging roads for people who actually had to flee. After the Maui fires, officials worried that families were turning down help because they feared a colonial land grab. And while we’d all like to blame bots and foreign trolls, the truth is more uncomfortable: many of the people spreading false rumors genuinely believe them. One major source of Maui conspiracy content was an account called Cultivate Elevate, run by a health food entrepreneur with more than half a million followers, who posted video after video claiming the fires had been deliberately set. Whether it’s mental shortcuts, confirmation bias, or just the human need to make sense of senseless tragedy, ordinary people are both the carriers and the victims of this plague.
The newest wildcard is artificial intelligence. Chatbots like ChatGPT and Gemini are now part of everyday life for about a quarter of American adults, and searching for information is their most common use. On the surface, that might sound promising—these models often draw on credible news sources and official websites, and they can debunk myths better than a random social media post. Ask ChatGPT about “Maui directed energy weapon” and it will tell you, quite firmly, that there is no credible evidence for that theory. But in an emergency, can we really trust a machine to know what’s happening right now? A model might pull from an article that’s days old, missing the fact that a storm has changed course or an evacuation order has just been issued. It might give you a confident answer about your eligibility for FEMA aid that’s completely wrong. Researchers are only beginning to study how AI operates in disaster situations, but the uncertainty is real. The Red Cross has built a chatbot called Clara to help people navigate emergencies. MIT has created something called CrisisChat. The Trump administration’s America.gov offers information about FEMA assistance. But as one crisis expert pointed out, the question isn’t just whether AI is accurate—it’s whether people trust it. If people trust it, they’ll act on it, whether it’s true or not. And right now, we don’t really know how people in disaster zones are using these tools, or whether they’re helping or making things worse.
So what can actually be done? Experts agree that you can’t fight misinformation in the moment. You have to build trust before the storm hits. FEMA began experimenting with this toward the end of the Biden administration. In Maui, the agency partnered with a Native Hawaiian employee who used his own networks to reassure people that no one was coming to take their homes. It worked with the Rosalynn Carter Institute for Caregivers to push disaster information through home health aides and other trusted figures. The idea is simple: people may not trust the government, but they trust their neighbor Jim, the firefighter who posts on Facebook. “From a government perspective, they’re just like, ‘They don’t know us,’” said one communications expert. “But they don’t have to know us, they know Jim down at the fire station.” The catch is that this only works if you lay the groundwork before disaster strikes—and even then, some people will always choose the conspiracy over the weatherman. There’s no decisive victory to be won, no final debunk that ends the war. As one former FEMA official put it, fighting misinformation in the modern era is less like swordplay and more like trench warfare. You slog forward, you lose ground, you dig in, and you try to make sure that in the moment when it matters most, enough people hear the truth to keep themselves and their neighbors alive. It’s an uphill battle, and the stakes have never been higher—but for the people on the front lines of climate disaster, it’s the only battle that counts.