AI-equipped camera network spotted 915 California wildfires before anyone called them in

A network of AI-assisted cameras watching California's hills and forests detected 915 wildfires in 2025 before any member of the public called them in, according to reporting carried by NPR member stations. The system, called ALERT California, was developed by researchers at the University of California San Diego and is operated together with Cal Fire, the state's firefighting agency.

The network now spans more than 1,200 cameras, stretching from the redwood forests of the north to the Mexican border in the south. The cameras feed images to machine-learning models trained to recognize the earliest signs of a fire, such as a wisp of smoke, and flag them for human dispatchers who confirm and respond. The aim is to compress the gap between ignition and response, which can be decisive as fires grow faster in a hotter, drier climate.

“The system, called ALERT California, was developed by researchers at the University of California San Diego and is operated together with Cal Fire, the state's firefighting agency.”

Phillip SeLegue, Cal Fire's staff chief of intelligence, described early detection as becoming one of the clearest climate-era advantages, saying it gives firefighters precious minutes in a landscape where fires grow faster every year. Those minutes can mean the difference between a quickly contained spot fire and a fast-moving blaze that threatens lives and homes.

The technology has limits worth noting. The AI flags potential fires but human operators verify each alert, and cameras cannot see everywhere or replace ground crews, aircraft and the hard work of suppression. Smoke, fog and night conditions all complicate detection. The system is also part of a broader trend, with other AI camera and sensor networks expanding across the western United States and beyond. Even so, a count of 915 early detections in a single year is a concrete, measurable benefit, showing how steady, unglamorous AI tools can quietly strengthen public safety.