August 17, 2026 - Fatal Waymo Crash Raises Accountability Questions

This is the Telemetry Transportation Daily for August 17, 2026, and I'm Sam Abuelsamid, Vice President of Market Research for Telemetry. 

Over the past decade, there have been numerous fatal crashes involving vehicles operating with various levels of driving automation. Most of those have involved Tesla vehicles, which are technically still Level 2, even for the so-called Full Self-Driving, meaning that the human driver must be supervising the automation and ready to take over at any time. Ultimately, the human driver is responsible for anything that happens leading to a crash. 

However, when we shift to higher levels of automation, things get more complicated, as there may not be anyone in the vehicle, and if they are simply a passenger, they can't be held responsible for the crash. The first known fatal crash with an automated vehicle occurred with an Uber prototype in 2018, and the on-board safety operator eventually pleaded guilty to a charge of endangerment and served three years of probation.  In October 2023, a pedestrian in San Francisco was struck by a human-driven vehicle, thrown into the path of a Cruise robotaxi, and then dragged. The victim was severely injured and survived, but Cruise never really recovered from withholding information about what happened and was ultimately shut down. 

Recently, a similar incident took place in Dallas where an SUV struck a pedestrian, throwing them into the side of an unoccupied Waymo robotaxi, and that victim died. While many details about exactly what happened are not yet publicly known, we don't know how much data from the automated vehicle Waymo has shared with law enforcement. While the Waymo clearly didn't cause the crash, we don't know if the actions of the vehicle may have worsened the situation or led to the victim's death. 

It's critically important that as automated vehicle operations start to scale, we put policies in place that require prompt sharing of all data from AVs whenever there is a crash. We also need to develop policies for how the companies that built automated driving systems will be held responsible when the inevitable crashes do happen. Human drivers can be held criminally responsible for actions leading to crashes. We don't yet have a robust legal framework for automated systems, whether they are vehicles or any other type of automation, such as AI, apart from civil product liability. That's just not good enough.

Thanks for listening.