A crash involving a Waymo robotaxi—or any fully driverless vehicle operating under California’s autonomous-vehicle rules—raises questions that a normal fender-bender does not. Who was “driving”? Was the autonomous system engaged? Is the claim about software and sensors, fleet operations, another human driver, or some mix of all three?

California has not replaced negligence and products law with a special robotaxi damages code. It has layered DMV permitting and financial-responsibility rules on top of ordinary civil liability. This article explains the main liability themes in plain English. It does not invent crash statistics, safety rankings, or outcome promises. For related practice context, see our AI car accident attorney page and car accident lawyer overview.

Autonomous vehicle vs. driver-assist: start with the definition

Vehicle Code section 38750 defines “autonomous technology” as technology capable of driving a vehicle without the active physical control or monitoring by a human operator. An “autonomous vehicle,” for purposes of that division, is a vehicle with integrated autonomous technology meeting SAE Levels 3, 4, or 5 as referenced in the statute.

Just as important is what the statute says an autonomous vehicle is not. Vehicles with collision-avoidance aids—blind-spot assist, automated emergency braking, adaptive cruise, lane keep, and similar systems that still require a human’s active control or monitoring—are carved out of that definition. Marketing labels are not the legal test. Whether a system was engaged, and what capability it actually had at the moment of impact, are fact questions.

Robotaxi operations of the sort associated with companies like Waymo typically involve vehicles designed to operate without a conventional human driver in the loop for parts of a trip. That operational model shifts the liability conversation toward the manufacturer, the fleet operator, software and sensor performance, and remote-operations practices—while still leaving room for claims against other road users who were careless.

What Vehicle Code § 38750 requires (and what it does not decide)

Section 38750 is primarily a regulatory and permitting statute. Among other things, it:

  • Requires manufacturers testing or deploying autonomous vehicles to show substantial financial responsibility—an instrument of insurance, surety bond, or self-insurance in the amount of $5,000,000—and to provide evidence to the DMV as regulations require
  • Sets certification and safety-feature expectations before public-road operation beyond testing rules
  • Requires autonomous technology to capture and store sensor data for at least 30 seconds before a collision while operating in autonomous mode, in a read-only format, preserved for three years after the collision date (see § 38750(c)(1)(G))

Those rules matter for investigation and for confirming that a manufacturer met permitting conditions. They do not by themselves tell a jury who was at fault in a civil case. DMV collision reporting and NHTSA inquiries may illuminate what happened; civil fault is still decided under tort and products principles on the facts of the crash.

Liability themes after a robotaxi crash

1. Other human drivers or road users. Many AV collisions still involve a conventional driver who rear-ends a stopped robotaxi, runs a light, or otherwise fails to use ordinary care. Civil Code section 1714 remains the backbone negligence rule. Comparative fault can allocate responsibility among multiple actors.

2. Company / operator negligence. Fleet operators and manufacturers may face negligence theories tied to how vehicles are dispatched, maintained, mapped, remotely supervised, or returned to service after known issues. The precise defendant names depend on corporate structure and contracts—parent companies, operating entities, and technology suppliers are not always the same.

3. Product liability. California recognizes strict products liability for defective products that cause injury. Foundational cases include Greenman v. Yuba Power Products, Inc. (1963) 59 Cal.2d 57, with design-defect analysis developed in cases such as Barker v. Lull Engineering Co. (1978) 20 Cal.3d 413. In an AV setting, defect theories may focus on software decision-making, sensor fusion, perception failures, inadequate fail-safe behavior, or insufficient warnings and instructions—subject to proof on the specific vehicle and build.

4. Mixed scenarios. A robotaxi may behave unexpectedly and another driver may also be careless. California comparative fault principles can assign percentages; product and negligence theories can travel in the same case when the facts support both.

No single theme “always” wins. The engaged status of the autonomous system, event data, video, map versioning, and the other parties’ conduct all matter.

Evidence that often decides these cases

Preserve early:

  • Police or CHP report numbers and any body-worn or dash video references
  • Photos and video of the scene, vehicle positions, and any AV exterior markers or trade dress
  • Witness information
  • Your medical records and a clear history of symptoms
  • Any communications from the AV company or its insurers
  • Your own vehicle’s event data recorder information, if available, through proper channels

On the AV side, section 38750’s pre-crash sensor-data retention requirement is one statutory hook for technical evidence when the vehicle was in autonomous mode. Companies may also hold telemetry, remote-assistance logs, and camera footage. Formal preservation letters and discovery tools are often needed; informal requests are not always enough.

Do not rely on social-media summaries of “what Waymo did.” Rely on authenticated data from the vehicles and systems involved in your crash.

Insurance and claims handling

AV manufacturers’ financial-responsibility filings under section 38750 sit alongside ordinary auto liability and first-party coverages that may apply depending on who was hurt and which vehicles collided. Passengers, pedestrians, cyclists, and other motorists can face different claim paths. Expect multiple carriers, complex correspondence, and requests for recorded statements. You are not required to guess about software architecture in an early call with an adjuster.

If your own insurer mishandles first-party benefits, that raises separate issues from the AV defendant’s liability. Deadlines for filing suit against private parties are generally two years under Code of Civil Procedure section 335.1, with shorter government-claim rules when a public entity is involved.

What not to assume

  • That “no human driver” means automatic company liability in every fact pattern
  • That a DMV report equals a civil verdict
  • That driver-assist branding and true autonomous operation are the same thing
  • That early settlement offers reflect the full technical picture

Robotaxi litigation is evidence-heavy. The confident approach is to lock down the data, identify every potentially responsible party, and apply ordinary California tort and products principles to uncommon facts.

Advo Law, APC evaluates motor vehicle and emerging-technology crash claims statewide on a contingency fee basis, with offices in Glendale and La Cañada Flintridge. Attorney Hov J. Kachatoorian and the firm focus on trial-ready preparation rather than volume advertising slogans.

Contact Advo Law

If you were injured in a crash involving a Waymo vehicle or another autonomous or robotaxi system in California, get advice before evidence disappears.

Advo Law, APC
Glendale: 111 E. Broadway, Suite 210, Glendale, CA 91205
La Cañada Flintridge: 1433 Foothill Blvd., Suite 207, La Cañada Flintridge, CA 91011
Phone: (800) 808-4613
Contact us · AI / autonomous vehicle accidents · Car accident lawyer

Contingency fee representation. Serving clients throughout California.

Attorney advertising. This article is for general information only and is not legal advice. Reading it does not create an attorney-client relationship. Autonomous-vehicle regulations and case law continue to develop; consult counsel about your specific situation.