The first-ever fatal crash involving a fully autonomous vehicle took place in San Francisco on January 19. The driverless car belonged to Waymo, but the tragic accident was not caused by its technology. A human-driven SUV, traveling at high speed, triggered a six-car pileup that left one person dead and five others injured. Even a pet dog lost its life in the collision.
This incident, along with other recent crashes involving Waymo’s self-driving cars, has intensified discussions about the safety of autonomous vehicles. However, a closer look at Waymo’s crash history suggests that most of these accidents are caused by human error rather than flaws in the technology. When compared to traditional human drivers, Waymo vehicles appear to be significantly safer.
The fatal January crash happened when a Waymo vehicle, with no driver or passengers, stopped at a red light. A car pulled up behind it, only for an SUV to rear-end the stopped vehicles at high speed. The force of the collision resulted in a devastating pileup.
A similar incident occurred in October in San Francisco. A Waymo vehicle, again stopped at a red light, was struck when another car crossed the centerline and crashed into an SUV that was waiting beside it. The impact pushed the SUV into the Waymo, causing significant damage and leaving one person seriously injured.
These accidents, while severe, followed a familiar pattern seen in Waymo’s crash reports. The company’s autonomous cars tend to be involved in accidents where they are simply obeying traffic laws, while human drivers make reckless mistakes such as speeding, running red lights, or losing control of their vehicles.
Between July 2024 and February 2025, Waymo reported 38 crashes serious enough to either trigger an airbag or result in injuries. A detailed review of these incidents reveals that the vast majority were not caused by the self-driving system.
Sixteen of these crashes involved another vehicle striking a stationary Waymo, with ten of those being rear-end collisions. Eight additional accidents occurred when a moving Waymo was rear-ended by another car. Five crashes happened when another driver veered into Waymo’s right of way, often by running a red light or a stop sign. In three separate cases, passengers of a Waymo vehicle accidentally opened the car door into passing traffic, hitting a car or bicycle despite Waymo’s built-in “Safe Exit” warning system.
Two reported cases appeared to involve no crash at all. In one incident, a human driver, startled by a Waymo’s lane adjustment, overreacted and crashed into a curb. In another, a pedestrian claimed to have been injured after approaching a stopped Waymo and making possible contact with its exterior.
Waymo was clearly at fault in just one case, when a self-driving vehicle in Los Angeles hit a plastic crate and pushed it into the path of an oncoming scooter, causing the rider to fall. There were three other incidents where Waymo’s role in the crash was unclear, involving situations where another vehicle attempted an unprotected left turn in front of a Waymo. Even if all of these cases were blamed on the autonomous system, the data still shows that the majority of crashes were caused by human drivers.
When comparing per-mile crash rates, Waymo vehicles consistently perform better than human-driven cars. Over 44 million miles of operation in San Francisco and Phoenix, Waymo vehicles were involved in only 13 crashes serious enough to deploy an airbag. Based on accident statistics from human-driven cars in the same locations, an equivalent number of human-driven miles would have resulted in 78 such crashes, making Waymo’s system 83% safer in this regard.
In terms of injury-causing accidents, human drivers would have been expected to cause 190 crashes over 44 million miles, whereas Waymo reported only 36, an 81% reduction. Waymo’s safety improvements have even increased over time. In September 2024, the company’s data suggested a 73% reduction in injury-causing crashes. By the end of the year, that number had improved significantly.
Insurance data further supports the claim that Waymo’s vehicles are safer than human-driven ones. Between January and July 2024, Waymo cars covered 25 million miles and faced only two pending injury claims. If the same distance had been driven by human-operated cars, insurance estimates suggest there would have been 26 bodily injury claims. Even if both pending claims against Waymo succeed, that still represents over a 90% reduction in insurance claims for bodily injury.
The difference in property damage claims is almost as significant. Over 25 million miles, Waymo vehicles generated just nine successful or pending claims, while human-driven vehicles would have been expected to produce 78 claims—a reduction of 88%.
