AI-Generated Wildfire Images Spread Emergency Misinformation During British Columbia's 2025 Fire Season

During British Columbia’s 2025 wildfire season, AI-generated images depicting wildfire scenes circulated widely on social media platforms, prompting an official warning from the BC Wildfire Service on August 5, 2025.

The service identified multiple fabricated images being shared on social media that inaccurately portrayed fire conditions around British Columbia. One image was posted by a self-described “digital creator” on Facebook on July 31 with a caption referencing the Drought Hill fire near Peachland. The following day, the caption was edited to add a disclaimer that the image was AI-generated and intended for “illustrative purposes only” — but by then it had already been shared as authentic documentation of the fire.

The BC Wildfire Service noted that many of the AI-generated images exaggerated the size and intensity of blazes burning around the province, stoking fear. However, the service also warned of the inverse risk: someone could generate an image showing an aggressive wildfire behaving with less intensity, which could mean someone in danger pays less attention. Both directions of misinformation — exaggeration and minimization — carry safety consequences in an emergency where people make evacuation decisions based on perceived fire behavior.

The service emphasized that people routinely turn to social media for wildfire updates, and that the proliferation of AI-generated imagery “is a new wrinkle that could change someone’s decision-making in an emergency if they don’t know any better.” Whether well-intentioned or deliberately misleading, false or outdated information can make wildfire season worse, even causing people to take unnecessary risks.

The incident occurred during an active fire season with significant fire activity across BC, when accurate real-time information was critical for public safety. No specific injuries or deaths have been attributed to AI-generated wildfire misinformation, but the documented potential for AI imagery to affect emergency decision-making represents a novel harm pathway that had not previously been observed in the Canadian context.