Against a backdrop of increasingly frequent and severe extreme weather events globally, a new digital threat has emerged: artificially generated fake disaster footage that is spreading at alarming speed across online platforms in China, triggering tangible real-world disruptions, a new BBC analysis has found.
As climate change drives more intense and frequent natural disasters across East Asia, bad actors have seized on generative artificial intelligence tools to fabricate seemingly authentic footage of catastrophic events including floods, wildfires and landslides. These manipulated or fully AI-generated clips quickly go viral on Chinese social media platforms, where users hungry for real-time updates on developing disaster situations share them widely before fact-checking can take place.
The BBC’s analysis of multiple viral clips that circulated during recent extreme weather events in China highlights how easy it has become for bad actors to create convincing fake disaster content using off-the-shelf generative AI tools. Unlike traditional edited video fakes, modern AI-generated footage can mimic camera shake, on-the-ground reporter audio, and the chaotic visual texture of real breaking disaster coverage, making it extremely difficult for ordinary users to distinguish from authentic footage.
The rapid spread of these fake videos carries significant real-world consequences. In some cases, false footage has sparked unnecessary public panic, leading to runs on grocery stores in unaffected regions and overwhelming emergency response phone lines with false reports of danger. Local disaster management officials have been forced to divert time and resources away from actual response efforts to debunk viral fake content, creating additional strains during already challenging emergency situations.
This trend also reflects a broader global challenge: as extreme weather events become more common, the spread of AI-fabricated disaster misinformation has become a growing risk to public safety and emergency management worldwide. Experts quoted in the BBC analysis note that platform providers and regulators across the globe are still playing catch-up to develop effective systems to detect and remove AI-generated fake disaster content before it can spread widely and cause harm.