CA DMV releases incident reports for self driving car prototypes: 6 accidents, most in self-driving mode, caused by other drivers, no serious injuries
California reveals details of self-driving car accidents — LOS ANGELES (AP) — California state officials released reports Thursday …
Context & Ripple Effects
The disclosure formalizes what Google's program had been saying informally: weeks earlier, Google's self-driving car head reported just 11 minor accidents across 1.7 million test miles, none attributed to Google itself. By publishing per-incident reports rather than summary claims, the CA DMV turned those assertions into auditable public records.
It also set the template for the reporting regime that followed — annual disengagement tallies from Waymo and GM and cumulative accident analyses now flow through the same channel.
First-order effects
- Test operators like Google gain a standardized disclosure duty: every prototype collision on California roads becomes a public document, making the 'not our fault' framing independently checkable against police and witness accounts.
Second-order effects
- Because most of the six crashes were rear-endings by human drivers, the burden shifts to conventional motorists and their insurers — autonomous programs can point to state-filed reports when disputing liability, pressuring insurers to price human error around AVs differently.
Third-order effects
- Mandatory incident reporting becomes the regulatory backbone of AV testing: later analyses built on these filings, including [[a:932894|the review of all 88 California AV accidents from 2014–2018 finding autonomous vehicles at fault in just seven]], show how disclosure requirements compound into a usable safety dataset — and into Apple's own first-collision filing years later, reported under the same rules.
The trend: Autonomous vehicle oversight is converging on mandatory state-level incident disclosure, turning individual crash reports into the longitudinal safety record regulators and competitors alike rely on.