Los Angeles Is Burning—And AI Deepfakes Are Fueling Misinformation

As firefighters continue to battle wildfires across Los Angeles, another scourge threatens recovery efforts: AI-generated deepfakes.

On Wednesday, images began circulating on X (formerly Twitter) showing the iconic Hollywood sign engulfed in flames as fire raged across Mount Lee. However, these images were entirely fabricated using generative artificial intelligence tools.

Following the Hollywood sign images, other fabricated visuals surfaced online, including false depictions of looting—further adding to the confusion.

The wildfire and looting deepfakes are part of a growing trend of misinformation and conspiracy theories that spread during crises. Similar tactics were seen during past disasters, such as Hurricane Helene last September, when scammers disseminated AI-generated images of destruction in areas untouched by the storm.

“It's likely trolling, and they think it's funny,” Tim Weninger, professor of computer science and engineering at University of Notre Dame, told Decrypt. “It could also be social or political, like implying California deserves to burn, criticizing Governor Newsom, or reacting to DEI in firefighting. These are the main reasons, but there could be others.”

On Thursday, U.S. President-elect Donald Trump again posted on Truth Social to take aim at California Governor Gavin Newsom and the handling of the state’s water. Trump called him “Governor Gavin Newscum” after previously calling on the government to resign over his handling of the fires in Southern California.

Thanks to the connectivity of social media, these AI-generated images can spread rapidly. Unsuspecting users, particularly those unfamiliar with the area or crisis specifics, are particularly susceptible to believing that AI deepfakes are real. Compounding this issue is a lack of communication by government officials, which can make it harder for the public to distinguish fact from fiction.

Beyond inciting panic during crises, AI deepfakes are increasingly weaponized for political manipulation. Last year, as the East Coast of the United States grappled with the effects of Hurricane Milton and Helene, AI deepfakes began surfacing online showing destroyed buildings amid reports that areas predominately housing supporters of then-candidate Trump would not receive aid, further polarizing an already-tense political climate.

Weninger emphasized that the spread of misinformation often stems from societal values rather than technological limitations, noting that by sharing or liking content, individuals are also endorsing it.

“Social media problems are sometimes technology or fact-checking problems, but mostly they are values issues,” Weninger said. “As a society, we don’t value sharing accurate, truthful information enough. Every social media user must decide this for themselves. Technology can’t do it for them, nor can anyone else.”

A spokesperson with the Hollywood Sign Trust—a nonprofit organization that preserves, maintains, and promotes the Hollywood Sign—confirmed to Decrypt that the sign is, in fact, undamaged and standing.

“Griffith Park is closed for precautionary reasons,” they said. “The sign is safe and sound, and there is no validity to these false rumors.”

Edited by Andrew Hayward