If you were hurt in a crash involving a Tesla on Autopilot, a Waymo robotaxi, or any vehicle with lane-keep, adaptive cruise, or other automated features, you are dealing with a category of accident that most insurers, and many attorneys, are not fully equipped to handle. The technology moves faster than the law, and the tactics used to minimize your claim are already evolving to match.
We've recovered over $150,000,000 for injured people across Southern California. Here are the 10 questions you should be asking right now, with the legal framework behind each one.
At a glance:
- Driver-assist and fully autonomous vehicles carry different liability rules.
- Multiple defendants, including drivers, automakers, and software companies, can all share fault.
- Black box and telematics data can prove what the vehicle's system was doing before impact.
- California requires AV manufacturers to hold DMV permits and report crashes.
- A product liability claim may run alongside, and separate from, your auto negligence claim.
- Investigations take longer, but your two-year statute of limitations does not pause.
- Software update history is a key piece of evidence you have the right to demand.
- Autonomous rideshare crashes add fleet operators and software licensees to the defendant pool.
- Early settlement offers from insurers can erase your product liability claim if you sign.
- Evidence windows close fast. Getting legal help early is not optional in these cases.
1. The Difference Between Driver-Assist Features and Fully Autonomous Vehicles
Driver-assist technology and full autonomy are not the same thing, and the legal outcome of your crash depends heavily on which one was involved.
Driver-assist features include systems like Tesla's Autopilot and Full Self-Driving (supervised), GM's Super Cruise, Ford's BlueCruise, and similar lane-centering, adaptive cruise, and automatic emergency braking tools. These are classified as SAE Level 2 systems. The manufacturer's own documentation for every one of them states that a licensed human driver must remain attentive, hands-ready, and legally responsible at all times.
Fully autonomous vehicles are SAE Level 4 or Level 5 systems. They are designed to complete a trip without human intervention. Waymo's commercial robotaxis currently operating in Los Angeles are the most common example in Southern California. At this level, the technology, not the human passenger, is making the driving decisions.
The distinction shapes your entire claim. In a Level 2 crash, the driver who was not paying attention may be the primary liable party, and the automaker may be a secondary defendant if the system malfunctioned. In a Level 4 crash involving a commercial operator, the human in the vehicle may have no legal control at all, which shifts liability toward the company that deployed the vehicle.
In California, driver-assist systems still require a human driver to remain alert and in control, and that driver can be held legally responsible for a crash even when autopilot or a similar feature was engaged.
Takeaway: Before any other legal analysis, your attorney needs to determine exactly what system was active, at what automation level, and what that system's own documentation says the driver was supposed to be doing.
2. Who Can Be Held Responsible: The Driver, Automaker, or Software Company
California follows a pure comparative fault framework. Under Cal. Civ. Code §1714, every person is responsible for injury caused by their own negligence. In a multi-party automated vehicle crash, that rule does not eliminate any defendant simply because a computer was also involved.
The potential responsible parties include:
- The human driver or operator, if they failed to maintain the attention their vehicle's automation level required.
- The vehicle manufacturer, if a defect in the hardware, sensors, or software caused or contributed to the crash.
- The software developer, if the automated decision-making system contained a flaw that was separate from the vehicle's physical design.
- A component supplier, such as a LiDAR, radar, or camera vendor, if a sensor failure contributed to the crash.
- A fleet operator or rideshare company, if they deployed an autonomous vehicle that was not properly maintained or certified.
Proving which party bears what share of fault requires the kind of forensic analysis that most standard auto-accident firms do not have the infrastructure to pursue. This is why autonomous and driver-assist crashes are almost always more complex than a standard rear-end or intersection collision.
Takeaway: Do not assume the driver is the only defendant. A thorough investigation often reveals that the manufacturer or software vendor shares fault, and California law allows all of them to be named.
3. Why Black Box and Telematics Data Matter More in These Crashes
Telematics data from automated vehicles can be overwritten within days of a crash, making immediate preservation of that evidence one of the most time-sensitive actions after any autonomous-vehicle accident in California.
Every modern vehicle with driver-assist or autonomous features logs a continuous stream of data: speed, steering inputs, brake application, sensor readings, driver eye-tracking data (in some systems), hands-on-wheel detection, lane position, and system alerts. This is not a single "black box" chip. It is typically a distributed set of logs stored in the vehicle's primary computer, in manufacturer cloud servers, and on third-party telematics platforms.
Under California Evidence Code §1552, electronically stored information that is printed or stored in a reasonably reliable way is presumed authentic. That presumption helps you, but only if the data still exists. Manufacturers routinely push over-the-air software updates that can overwrite operational logs. Some logs cycle on a rolling window basis and delete themselves after a set number of hours or miles.
The moment you retain an attorney, a legal hold letter should go to the vehicle manufacturer, the rideshare operator if applicable, and any telematics third-party provider. This creates a documented, date-stamped obligation to preserve the data, and if the defendant allows it to be destroyed afterward, you may be entitled to a spoliation inference at trial.
Takeaway: Preserving telematics data is time-sensitive. An attorney should send a legal hold letter within days of the crash, not weeks.
4. How California Currently Treats Autopilot-Related Accident Claims
California Vehicle Code §38750 requires manufacturers testing fully autonomous vehicles on public roads to obtain a DMV permit and report every collision, creating an official record that can support your claim.
California is one of the most active jurisdictions in the country for autonomous vehicle regulation, and that creates paper trails that matter for your claim.
Under Cal. Veh. Code §38750, manufacturers who test autonomous vehicles on California public roads must obtain a permit from the DMV, and they must report every collision involving a permitted autonomous vehicle. These collision reports are public records. If your crash involved a permitted test vehicle or a commercially deployed autonomous vehicle, that report likely exists and belongs in your case file.
For marketed driver-assist features like Autopilot, the regulatory picture is less settled. There is no California statute that automatically assigns liability to Tesla when its Autopilot feature is engaged at the time of a crash. Instead, courts apply standard negligence and products-liability analysis. NHTSA has opened multiple investigations into Autopilot-related crashes and issued standing general orders requiring manufacturers to report Level 2 crashes involving airbag deployment or injuries. That federal investigation record can also be discoverable in a California civil case.
Takeaway: California's AV reporting requirements create an official incident record. Pull the DMV collision report and check for any open NHTSA investigations involving the specific system active in your crash.
5. When a Product Liability Claim May Run Alongside Your Auto Claim
A defective driver-assist feature can trigger a strict product liability claim against the vehicle manufacturer under California law, entirely separate from any negligence claim against the driver.
California's strict products liability doctrine, rooted in the California Supreme Court's decision in Greenman v. Yuba Power Products (1963) and codified in subsequent case law, holds that a manufacturer is liable for injuries caused by a product defect without requiring the injured person to prove the manufacturer was negligent. There are three defect theories available:
- Manufacturing defect: the specific vehicle was built incorrectly and deviated from its intended design.
- Design defect: the entire product line was designed in a way that was unnecessarily dangerous.
- Failure to warn: the manufacturer did not adequately inform users of the risks of operating the automated feature.
In a driver-assist crash, any one of these theories can apply. A software bug that caused the system to misread lane markings could be a design defect. An inadequate alert system that failed to warn the driver to take over could be a failure to warn. A sensor that was installed with a manufacturing flaw in that specific vehicle could be a manufacturing defect.
This product liability claim runs alongside, not instead of, your negligence claim against the driver. Both can be filed in the same lawsuit. If you settle with the driver's insurer and release all claims without first understanding whether a product defect contributed to the crash, you may inadvertently extinguish your right to sue the manufacturer.
Takeaway: Never sign a release without understanding whether a product liability claim also exists. The two claims are legally independent, and settling one does not automatically settle the other.
6. Why These Investigations Often Take Longer Than Typical Crash Claims
California's personal injury statute of limitations is two years from the date of injury under Code of Civil Procedure §335.1, and it continues to run while investigations, data analysis, and regulatory reviews unfold.
A standard two-car collision claim can be investigated, evaluated, and resolved in a matter of months. An autonomous or driver-assist crash claim rarely follows that timeline.
The reasons include:
- Forensic data analysis requires specialized automotive engineers, not just accident reconstructionists.
- Proprietary software review often involves a fight over discovery. Manufacturers routinely assert trade secret protection over their algorithms and resist disclosure.
- Regulatory timelines from NHTSA or the NTSB can take years, and their findings, while not binding on a California civil court, carry significant persuasive weight.
- Multiple defendants mean multiple insurers, multiple defense attorneys, and extended negotiation or litigation timelines.
- Expert witness development in a novel technology field takes time that standard PI cases do not require.
Under Cal. Code Civ. Proc. §335.1, you have two years from the date of injury to file a personal injury lawsuit in California. If a government-owned autonomous vehicle is involved, Cal. Gov. Code §911.2 requires you to file a government tort claim within six months of the incident. Neither deadline pauses while the investigation proceeds.
Takeaway: The longer the investigation, the more pressure on your filing deadline. Start the legal process early, even if you expect the investigation to take time.
7. What to Ask About the Vehicle's Software and Update History
Modern driver-assist vehicles receive over-the-air software updates the same way a smartphone receives app updates, sometimes several times per month. The version of software running on the vehicle at the moment of your crash may be different from the version running the week before, and different again from the version running today.
This matters for several reasons:
- The software version determines exactly what capabilities and limitations the system had at the time of the crash.
- A manufacturer may have released an update before the crash that addressed a known defect, raising the question of whether the vehicle had received it.
- A manufacturer may have released an update after the crash that altered the vehicle's behavior, which can affect the integrity of any post-crash testing or inspection.
You, or your attorney, have the right to demand:
- The software version log for the vehicle, showing every update and its timestamp.
- All manufacturer technical service bulletins or recall notices relating to the automated feature.
- Any internal communications about known defects in that software version, which may become discoverable through litigation.
The California DMV AV division also publishes manufacturer collision reports and disengagement reports (reports describing when an autonomous system handed control back to the human driver unexpectedly). Those public filings can establish a pattern of system failures before your crash.
Takeaway: Software version at the time of the crash, update history, and recall records are non-negotiable elements of any autonomous or driver-assist claim file.
8. How Autonomous Rideshare Vehicles Complicate Fault Questions
When a robotaxi or autonomous rideshare vehicle is involved, potential defendants can include the rideshare company, the fleet operator, the vehicle manufacturer, and the software developer, all at the same time.
When a privately owned Tesla driven by an individual is involved in an Autopilot crash, the defendant pool is relatively defined. When a commercially operated autonomous rideshare vehicle is involved, the structure changes significantly.
Under Cal. Pub. Util. Code §5430 et seq., Transportation Network Companies operating in California must maintain minimum insurance coverage. For autonomous vehicles without a safety driver, the coverage obligations during the active ride period generally attach to the TNC and the fleet operator. But coverage disputes arise quickly when automation is a contributing cause, because each insurer will argue that the other's coverage is primary.
Potential defendants in an autonomous rideshare crash include:
- The TNC (e.g., Waymo, Cruise, Uber if using AV technology), as the company responsible for deploying the vehicle.
- The fleet operator or contractor responsible for vehicle maintenance and readiness.
- The vehicle manufacturer, on a product liability theory.
- The software developer, if the autonomy stack is licensed from a third party.
If you were a passenger in the autonomous vehicle, you were likely doing nothing wrong, and the claim belongs entirely to the companies that put that technology on the road. If you were in another vehicle struck by the autonomous car, the same defendants are available to you.
Takeaway: Autonomous rideshare crashes require identifying the TNC, fleet operator, manufacturer, and software vendor as potential defendants before any coverage analysis can accurately assess available insurance limits.
9. Common Insurance Tactics Used in Tech-Related Crash Claims
Insurance carriers have already developed playbooks for driver-assist and autonomous vehicle crash claims. Knowing these tactics before you receive your first call from an adjuster puts you in a stronger position.
The most common tactics include:
Blaming the human driver. Because every Level 2 system requires driver attention, insurers routinely argue that the human failed to maintain proper control regardless of what the software was doing. This is often their opening position even when data exists showing the system issued no takeover alert.
Claiming data is proprietary. Manufacturers and their insurers frequently object to production of telematics and log data by asserting trade secret protection. These objections can be fought in California discovery proceedings, but they take time and legal effort.
Moving to settle before the investigation is complete. A quick offer that covers your medical bills and a small pain-and-suffering amount may feel fair before you know whether a product defect contributed to the crash. Accepting it and signing a general release extinguishes every claim, including a product liability claim worth many times more.
Disputing the injury's connection to the crash. In high-speed or complex crashes involving autonomous systems, insurers sometimes argue that a delayed software response, not the crash itself, caused the biomechanical forces that injured you. Expert medical and engineering testimony is required to rebut this.
Under California's Fair Claims Settlement Practices Regulations, insurers have defined timelines for acknowledging, investigating, and resolving claims. An attorney tracking those deadlines can hold the carrier accountable when delays appear strategic rather than procedural.
Takeaway: Do not accept a first offer, or any offer, in a driver-assist or autonomous crash until the full scope of potential defendants and claims has been investigated.
10. Why Getting Guidance Early Matters in This Evolving Area of Law
This area of law is not settled. California statutes, DMV regulations for autonomous vehicles, and federal NHTSA guidelines are all actively changing. Courts are still writing the decisions that will define how comparative fault applies when an SAE Level 3 or Level 4 system makes a driving error. The attorneys who are most effective in these cases are the ones who engaged before the evidence was gone.
The practical reasons to act quickly are:
- Telematics data has a short preservation window.
- Software updates can change the vehicle's behavior before forensic inspection occurs.
- DMV autonomous vehicle collision reports are filed quickly and create timelines.
- Multiple defendants mean multiple statutes of limitations may apply, some shorter than the standard two-year personal injury window.
- The legal framework is evolving, and an attorney tracking current NHTSA guidance and DMV rulemaking can anticipate how courts are likely to rule.
If your crash involved a vehicle with any automated feature, whether it was in a fully autonomous robotaxi or a standard car with a lane-keep assist system, the questions above have to be answered before your claim can be accurately valued. We handle personal injury cases that involve this level of technical complexity, and we have recovered over $150,000,000 for injured people across Southern California.
If you or someone you know was hurt in a self-driving or driver-assist crash, call (818) 794-9947 for a free review of your case. No fee unless we win.
Frequently Asked Questions
Who is liable in a self-driving car accident in California?
Liability depends on the automation level involved. In a Level 2 driver-assist crash, the human driver, the vehicle manufacturer, a software developer, or a component supplier can all be liable, and California's comparative fault rules allow multiple defendants to share responsibility. In a Level 4 autonomous vehicle crash, the commercial operator, fleet manager, and manufacturer are typically the primary defendants because no human was required to be in control.
Does Tesla Autopilot change who is at fault in a crash?
Tesla Autopilot is a Level 2 driver-assist system. Tesla's own documentation requires the driver to remain attentive and ready to take control at all times. California courts apply standard negligence rules, meaning the driver who failed to maintain attention can still be held liable. However, if a software defect contributed to the crash, a product liability claim against Tesla may also exist and can be filed alongside the negligence claim.
Can you sue the manufacturer after a driver-assist accident?
Yes. Under California's strict product liability doctrine, if a defect in the vehicle's hardware, software, or design contributed to the crash, you may sue the manufacturer without having to prove they were negligent. There are three defect theories available: manufacturing defect, design defect, and failure to warn. This claim is independent of any negligence claim against the driver, and both can be pursued in the same lawsuit.
How is data used in autonomous vehicle accident claims?
Telematics, sensor logs, camera footage, and control-input data generated by the vehicle can prove what the automated system was doing in the seconds before impact, whether a takeover alert was issued, and whether the driver had hands on the wheel. Under California Evidence Code §1552, electronically stored information is presumed authentic. Preserving this data quickly is critical because it can be overwritten by manufacturer updates or cycling log storage within days of the crash.
What is the deadline to file a lawsuit after a self-driving car accident in California?
California's standard personal injury statute of limitations is two years from the date of injury under Code of Civil Procedure §335.1. If a government entity or government-operated autonomous vehicle is involved, a government tort claim must be filed within six months of the incident under Government Code §911.2. These deadlines do not pause during investigations or regulatory reviews, so legal action should be initiated as early as possible.
Does California have specific laws governing autonomous vehicle accidents?
California Vehicle Code §38750 requires manufacturers testing autonomous vehicles on public roads to obtain a DMV permit and report every collision. Beyond testing requirements, California does not yet have a statute that automatically assigns liability to a manufacturer when an autonomous feature is involved in a crash. Courts apply existing negligence and product liability frameworks, while California DMV and NHTSA continue to develop their regulatory guidance.
What should I do immediately after a crash involving a driver-assist vehicle?
Document the scene thoroughly: photograph the vehicle, any visible sensor damage, road markings, and signage. Note the exact driving mode the vehicle was in if you can access that information. Do not allow the vehicle to be moved to a dealer for software updates before a legal hold is in place. Seek medical care, report the crash to the California DMV if an autonomous vehicle was involved, and contact an attorney before giving a recorded statement to any insurer.
Can I still recover damages if I was a passenger in an autonomous rideshare vehicle that crashed?
Yes. As a passenger, you were not operating the vehicle and bear no comparative fault for the crash. Potential defendants include the TNC or rideshare company, the fleet operator, the vehicle manufacturer, and the software developer. California's Public Utilities Code §5430 requires Transportation Network Companies to carry minimum insurance, and additional coverage may be available through the manufacturer's liability policies.
Reviewed by Minas Nordanyan, CA Bar No. 296806. Last updated September 2026. This article is for informational purposes only and does not constitute legal advice. Case outcomes depend on the specific facts of each matter. Past results do not guarantee future outcomes.