You remember the 2023 Maui wildfires, don’t you? More than a hundred people died, nearly two thousand homes were destroyed, and the historic town of Lahaina was all but erased. The official story was that a broken power pole owned by the local utility sparked dry grass on a windy day. But in the days that followed, a very different story exploded across social media. Thousands of people shared videos and posts claiming the fires had been started by a “direct energy weapon,” a laser beam from space, or a government plot to clear the land for a high-speed rail line or a “smart city.” Right-wing personalities amplified the claims, accounts linked to foreign influence operations pushed a “weather weapon” narrative, and rumors swirled that FEMA would seize the homes of anyone who applied for disaster aid. That last one was especially cruel in a place with deep Indigenous trauma around land theft. This is what a perfect storm of misinformation looks like. Extreme weather has always caused confusion, but never before has the information landscape been so fractured. Social media companies have abandoned fact-checking, government agencies are stretched thin, politicians distort disasters for gain, and generative AI has added a new layer of uncertainty. A recent report found that the vast majority of teens encounter conspiracy theories on social media, and most of them are inclined to believe what they see. As Amber Silver, a disaster communications expert at the University at Albany, put it, disasters are urgent and uncertain, and victims desperately need facts but have almost no reliable way to get them. All the conditions are perfectly aligned for misinformation to thrive.
One of the most striking examples of how this happens is the way an old, unrelated photo morphed into a global conspiracy. It started innocently enough in January 2018, when a Marathon Petroleum refinery in Canton, Ohio, flared gas during a mechanical issue. A local resident snapped a blurry picture, and the image included a lens flare pointing upward. That photo sat on Facebook for years. In May 2018, SpaceX posted an Instagram photo of a Falcon 9 rocket launch from California, which also produced a bright vertical streak. In May 2023, a windstorm in Santiago, Chile, caused an electrical transformer to explode, and a resident captured the event on video, complete with its own dramatic lens flare. Then came Maui in August 2023. Within hours, those old images and videos were being passed off as evidence of a secret energy weapon aimed at Lahaina. They were viewed and shared millions of times. One Hawaii-based Instagram user posted the old Ohio refinery photo and suggested the fires were an “attack.” When she realized her mistake, she made a follow-up video apologizing, saying it was important to vet information. But it was too late. A larger account, run by a health food entrepreneur with a huge following, had already taken a screenshot of her now-deleted post and claimed it showed a laser beam targeting the city. That video got more than 400,000 views, forty times more than the apology. Then came screenshots of screenshots of screenshots, each step further from the truth. This is how misinformation actually works: images and facts blend with old fears and suspicions, and a conspiracy takes on a life of its own.
The consequences of this chaos are not abstract. Bad information after a disaster can cost lives, slow recovery, and erode trust in the very people trying to help. One Department of Homeland Security study found that an inaccurate map tweeted during a 2017 California flood triggered a mass evacuation that didn’t need to happen. Another study found that an earthquake in Mexico caused an 11 percent drop in public trust in government. A review of recent research concluded that misinformation often leads to resource misallocation, overwhelmed emergency services, and diverted attention away from genuine needs. The most visible target is FEMA. Deanne Criswell, who led the agency under the Biden administration, said FEMA has been the subject of misinformation since its founding. The old myth that FEMA plans to lock disaster victims in “camps” has never really gone away. It feeds on a deep suspicion of government, and it makes people hesitate to ask for help when they need it most. During Hurricane Helene in 2024, a rumor that someone was coming to North Carolina to shoot FEMA workers forced Criswell to halt field operations for hours. She and her deputy chief of staff were both “swatted,” with anonymous callers sending SWAT teams to their homes. Meanwhile, right-wing accounts accused Biden of withholding aid, and Elon Musk claimed FEMA was blocking him from delivering supplies. His retraction got a million views, but his original accusation got thirty million. The damage was done. And mistrust doesn’t only come from conspiracy theorists. Undocumented people may fear deportation if they interact with officials. Communities with painful histories of policing may not trust authority figures. All of it leads to the same outcome: fewer people getting help, and more people putting themselves in danger.
Social media is the incubator for most of this misinformation. Rumors around disasters are nothing new. After the 1889 Johnstown flood, people blamed wealthy industrialists for causing the disaster. After Hurricane Katrina, some claimed FEMA was putting survivors in concentration camps. But Facebook, Instagram, and X allow rumors to circle the globe in minutes, and they make it almost impossible to trace where a falsehood began. Studies are beginning to quantify the problem. After a 2023 earthquake in Turkey, about a quarter of residents surveyed believed some version of the conspiracy that the United States had caused the quake with a seismic weapon. A report from the Center for Countering Digital Hate found that most conspiracy posts came from verified accounts, giving them an air of legitimacy. Even harmless fakes, like the famous doctored image of a shark swimming on a flooded highway, have fooled media outlets and even a U.S. senator. A meta-study published last year found that misleading messages often outpaced official communications, increasing anxiety and reducing trust. The consequences can be strange and dangerous. After a 2018 earthquake in Japan, communities with low trust in government evacuated even when they didn’t need to, clogging roads and putting real evacuees at risk. After Maui, Criswell worried that the fear of a colonial land grab caused families who needed help to reject it. Many people assume bots are the main problem, but research suggests that most people who share false rumors actually believe them. The account that amplified the Maui laser theory, for example, had more than half a million followers and was run by a real person with a real audience. It wasn’t just a foreign troll farm. It was ordinary people passing along something that felt true.
The rise of generative AI adds a new and poorly understood layer to this crisis. About a quarter of U.S. adults now use AI chatbots daily, and searching for information is one of the most common uses. Chatbots like ChatGPT and Gemini often draw on news articles and official sources, which can actually help debunk misinformation. Ask ChatGPT about the Maui “directed energy weapon,” and it will tell you there is no credible evidence for that theory. But that doesn’t mean AI is reliable in an emergency. What if a model pulls from an outdated article that says a storm is not evacuating, but conditions have changed overnight? Can you trust a bot to accurately explain your eligibility for FEMA benefits? The technology is developing so fast that researchers haven’t had time to study how it behaves in disaster situations. Still, the intersection is growing. The Red Cross has built a chatbot named Clara to help people navigate crises. MIT has created something similar called Crisischat. The Trump administration launched America.gov, which includes information about FEMA assistance. But when it comes to whether people actually trust these tools, we are flying blind. Anita Atwell Seate, a crisis communication expert at the University of Maryland, says accuracy alone isn’t enough. You also need to know whether people trust the information, because if they trust it, they will act on it regardless of whether it’s true. That is a deeply human problem, and no algorithm can solve it.
So what can be done? There is no clean fix for a problem this tangled. Efforts to debunk false information are always slower and less viral than the misinformation itself. But experts agree that the best way to fight disaster rumors is to build trust before the emergency happens. Emergency managers need to partner with people who already have credibility in their communities. In the aftermath of Maui, FEMA worked with a Native Hawaiian employee who was able to reassure his network that the government was not coming for their land. The agency also partnered with the Rosalynn Carter Institute for Caregivers to spread disaster information through home health aides and caregivers. The idea is simple: people may not trust a government agency, but they trust Jim down at the fire station. If Jim posts a warning on Facebook, people listen. As Seate put it, the government doesn’t have to be known personally; it just needs to work through people who are. Justin Knighten, a risk communication consultant who led FEMA’s external affairs team under Biden, says the level of distrust and fragmentation is unlike anything we’ve seen before. When trust is low, he said, you’re working at a deficit, constantly trying to get ahead of threats that haven’t even formed yet. Fighting misinformation is not a duel with a clear winner. It’s more like trench warfare, a slow, exhausting struggle to gain ground in the information environment. But the stakes could not be higher. In the chaos of a hurricane or wildfire, the difference between a rumor and the truth can be the difference between life and death. That is why this work matters, and why we can’t afford to give up on it.