Car Accident Case Evaluation
Liability for a driverless car crash in California may extend beyond the person riding in the vehicle. Depending on how the collision occurred, potentially responsible parties may include the autonomous vehicle manufacturer, the company that developed or integrated the automated-driving system, a component manufacturer, a robotaxi or fleet operator, a maintenance provider, another negligent driver, or several parties at the same time.
Determining responsibility requires more than asking who was behind the wheel. In a fully autonomous vehicle, there may be no human driver making the immediate steering, braking, or acceleration decision. The investigation may instead focus on what the vehicle detected, what its software decided, whether the system was operating within its authorized conditions, whether a component failed, and which company controlled the evidence needed to explain the crash.
This issue is increasingly relevant in Santa Monica and the Los Angeles area, where fully autonomous ride-hailing vehicles operate on public streets. California also substantially updated its autonomous-vehicle regulatory framework in 2026. As the technology develops, serious collision cases can combine traditional negligence principles with products liability, commercial fleet responsibility, insurance coverage, electronic evidence, and highly technical vehicle data.
California Law Distinguishes Driverless Vehicles From Driver-Assistance Systems
The words “self-driving,” “Autopilot,” and “driverless” are sometimes used interchangeably, but the legal and technical distinctions matter.
California Vehicle Code section 38750 defines an autonomous vehicle as a vehicle equipped with technology meeting SAE Level 3, Level 4, or Level 5 automation. The statute specifically distinguishes those vehicles from ordinary collision-avoidance and driver-assistance features such as adaptive cruise control, automated emergency braking, lane-keeping assistance, and similar systems that still require active human control or monitoring.
That distinction can affect the investigation. A true driverless robotaxi may place responsibility for the driving task primarily on an automated driving system. By contrast, a Level 2 driver-assistance system generally still requires a human driver to remain attentive and able to control the vehicle.
Therefore, one of the first questions after an automated-vehicle crash should be: What technology was actually controlling the vehicle immediately before the collision?
Several Parties May Be Responsible for a Driverless Car Crash
Traditional California car accident claims often begin with the conduct of the drivers involved. Driverless-vehicle cases require a broader investigation. Readers seeking an overview of ordinary motor-vehicle liability can review our guide to California car accident claims.
In an autonomous-vehicle collision, the potentially responsible parties may include the following.
The Autonomous Vehicle Manufacturer
A vehicle manufacturer may face liability when a defect in the vehicle or integrated autonomous-driving technology contributes to an injury. Depending on the evidence, a claim could involve a manufacturing defect, design defect, inadequate warning, negligent development, or another product-related theory.
California products-liability law does not make manufacturers automatically responsible whenever a product is involved in an accident. The injured person must establish the elements of the applicable legal theory, including a defect and causation where required. California’s 2026 Civil Jury Instructions on products liability provide the framework California courts use for several product-defect claims.
The Autonomous-Driving Technology Company
The company responsible for the automated-driving system may be separate from the company that manufactured the underlying vehicle. That distinction can become important when the collision involves perception software, object classification, path planning, sensor fusion, mapping, or another automated decision-making function.
Whether a particular technology company can be held liable under strict products-liability principles, negligence principles, or another theory depends on its role in designing, manufacturing, integrating, supplying, updating, or operating the system. Software-related liability can present particularly complicated questions because not every technology provider occupies the same legal position as a traditional product manufacturer.
A Robotaxi or Fleet Operator
A company operating an autonomous ride-hailing fleet may possess significant control over the vehicles even when nobody is physically driving them. Its responsibilities may include dispatching vehicles, defining service areas, maintaining the fleet, monitoring operations, implementing software updates, responding to vehicle problems, and providing remote assistance.
A fleet operator’s potential liability depends on what it did or failed to do and whether that conduct contributed to the collision.
Sensor or Component Manufacturers
Autonomous vehicles can use cameras, radar, lidar, positioning systems, processors, braking equipment, steering systems, and other hardware. If a defective component contributed to the collision, the component manufacturer or another entity in the product’s distribution chain may become relevant.
The key issue is not simply whether a component was present. Investigators need to determine whether it failed, whether that failure affected the vehicle’s behavior, and whether the failure caused or contributed to the crash.
Maintenance and Repair Providers
Autonomous-driving systems depend on properly functioning hardware. Cameras can become obstructed or misaligned. Sensors can be damaged. Tires, brakes, steering components, electrical systems, and other equipment still require inspection and maintenance.
If negligent maintenance, repair, calibration, or inspection contributed to the collision, a maintenance company or contractor may share responsibility.
Another Driver or Road User
The presence of an autonomous vehicle does not automatically mean the automated system caused the collision. Another motorist may have run a red light, made an unsafe lane change, followed too closely, driven while distracted, or otherwise contributed to the crash.
California follows comparative-fault principles. Accordingly, responsibility can be divided among multiple actors when more than one person’s or entity’s conduct contributes to an injury.
California Negligence Law Still Applies to Autonomous-Vehicle Cases
California Civil Code section 1714 establishes the general principle that a person is responsible for injuries caused by a failure to use ordinary care in managing their person or property, subject to applicable limitations.
In a driverless-car case, negligence analysis may therefore focus on conduct such as:
- Placing a vehicle into service under conditions it could not safely handle
- Failing to properly maintain or calibrate equipment
- Ignoring known safety problems
- Failing to appropriately respond to detected system failures
- Negligently installing or integrating autonomous technology
- Failing to adequately monitor fleet operations
- Operating outside applicable permit or operational requirements
Which of these issues matters depends on the facts. A collision caused by another driver may involve very different liability questions from a crash caused by an autonomous vehicle failing to recognize a stopped vehicle, pedestrian, traffic signal, or roadway condition.
The Vehicle’s Sensors and Software May Become Central Evidence
One of the most important lessons from autonomous-vehicle litigation is that photographs of damaged vehicles may tell only part of the story. The collision may have resulted from decisions made inside the automated-driving system seconds before impact.
Depending on the vehicle, potentially important evidence may include:
- Camera recordings
- Radar information
- Lidar data
- Vehicle event logs
- Speed and acceleration data
- Steering inputs
- Brake commands
- Object-detection and classification records
- Mapping information
- Location and geofence information
- System warnings and diagnostic alerts
- Software version and update history
- Remote-assistance communications
- Maintenance and calibration records
- Prior incident information
California law specifically recognizes the importance of autonomous-vehicle data. Vehicle Code section 38750 requires qualifying autonomous vehicles to have a mechanism that captures and stores autonomous-technology sensor data for at least 30 seconds before certain collisions. The statute also requires that specified data be preserved for three years after the collision.
That statutory requirement does not mean every potentially useful electronic record will automatically be available to an injured person. Different entities may possess different categories of data, and obtaining proprietary electronic evidence may require preservation requests, discovery, subpoenas, technical inspections, or litigation.
Preserving Autonomous-Vehicle Evidence Can Be Critical
Evidence preservation should begin early after a serious driverless-vehicle crash. The goal is not simply to preserve photographs. It may be necessary to identify every company that possesses relevant electronic information.
For example, the vehicle manufacturer may control sensor information while the autonomous-driving company controls software logs. A robotaxi operator may possess trip records. A remote-assistance provider may maintain communications. A maintenance contractor may hold calibration and repair records.
A preservation request may therefore identify the specific vehicle, trip, date, time, collision location, occupants, and incident, while requesting preservation of potentially relevant electronic records and physical evidence.
Physical evidence from the scene still matters. Our guide explaining what to do after a California car accident discusses photographs, witness information, medical attention, insurance information, and other immediate steps that may help preserve a claim.
Investigators Need to Determine What the Vehicle Perceived
Autonomous-driving systems must perceive their environment and decide how to respond. Depending on the system, cameras, radar, lidar, maps, and other data sources may contribute to that process.
A useful investigation may ask:
- Did the vehicle detect the pedestrian, bicycle, vehicle, traffic signal, or obstruction?
- How did the system classify what it detected?
- Did different sensors provide inconsistent information?
- Did the software recognize the conflict?
- Did the system brake or steer?
- When did it initiate that response?
- Did a sensor failure or warning occur?
- Did the vehicle request human or remote intervention?
- Was the vehicle operating within the conditions for which the system was designed?
These questions can help distinguish a crash caused by another road user from one involving perception failure, defective system design, improper operation, inadequate maintenance, or another technology-related issue.
Geofencing and Operational Limits May Affect the Investigation
Autonomous vehicles are not necessarily designed to operate everywhere under every condition. Their permitted or intended operational environment may be limited by geography, road type, speed, weather, construction conditions, or other factors.
The California DMV autonomous-vehicle program regulates autonomous testing and deployment and identifies different permit categories and operating requirements.
California adopted substantially revised autonomous-vehicle regulations on April 28, 2026. Those regulations expanded and updated the state’s oversight framework, including rules governing testing and deployment.
An investigation may therefore consider whether the vehicle was operating within its approved or intended operational conditions when the crash occurred. That inquiry can include location, road type, weather, construction activity, speed, software configuration, and applicable permit restrictions.
Human Takeover Can Still Matter in Partially Automated Vehicles
Not every crash involving automated technology is a driverless-car crash. In vehicles using driver-assistance systems that require human supervision, the driver may still have a duty to monitor the roadway and respond when necessary.
That creates a different set of questions:
- Was the system represented as requiring continuous driver supervision?
- Was the driver paying attention?
- Did the vehicle issue a takeover warning?
- How much time did the driver have to respond?
- Did the system unexpectedly disengage?
- Was the warning adequate under the circumstances?
Responsibility may potentially be divided between the human operator and another party if both contributed to the collision.
Government Crash Reports May Provide Another Source of Information
Automated-vehicle crashes may generate records beyond the ordinary police collision report.
The California DMV maintains an autonomous-vehicle incident reporting system and publishes collision information associated with regulated autonomous-vehicle operations.
At the federal level, the National Highway Traffic Safety Administration’s automated-driving crash reporting program requires specified manufacturers and operators to report qualifying crashes involving automated driving systems and certain Level 2 driver-assistance systems.
These government records may help identify prior incidents, system information, regulatory investigations, or additional sources of evidence. However, a government report does not automatically establish civil liability.
For ordinary law-enforcement records, our resource explaining how to obtain a California accident report explains how collision reports may be requested from the investigating agency.
Insurance Coverage May Come From More Than One Source
Autonomous-vehicle collisions can involve multiple layers of insurance or financial responsibility. The applicable coverage depends on the vehicle, operator, fleet structure, responsible parties, and actual policy language.
Potential sources may include commercial fleet insurance, manufacturer coverage, another driver’s liability policy, an injured person’s own automobile policy, or other applicable coverage.
Vehicle Code section 38750 also contains financial-responsibility requirements for manufacturers operating qualifying autonomous vehicles under California’s statutory framework, including specified $5 million insurance, bond, or self-insurance requirements in certain testing and deployment contexts.
If an uninsured or underinsured human driver also contributes to the crash, an injured person’s own policy may become relevant. Our guide to uninsured and underinsured motorist coverage in California explains how UM/UIM coverage may operate when another responsible driver lacks sufficient insurance.
Damages Depend on the Injury, Causation, and Available Evidence
A person injured by an autonomous vehicle may potentially pursue the same general categories of compensatory damages available in other California personal injury cases when the evidence and applicable law support them.
Potential economic damages may include:
- Past medical expenses
- Reasonably necessary future medical care
- Lost income
- Loss of future earning capacity
- Rehabilitation expenses
- Property damage
- Other qualifying financial losses
Potential non-economic damages may include physical pain, emotional distress, inconvenience, loss of enjoyment of life, and other legally compensable harm.
The available amount cannot be determined simply from the fact that a driverless vehicle was involved. Liability, injury severity, medical causation, comparative fault, insurance, future consequences, and supporting documentation can all affect the claim.
Driverless Car Crashes in Santa Monica Require a Technology-Focused Investigation
Autonomous ride-hailing is no longer theoretical in Santa Monica. Driverless vehicles operate in the Los Angeles service area, including Santa Monica. That means a collision on a Santa Monica street may require both an ordinary traffic investigation and a separate examination of autonomous-driving technology.
For example, a crash near a busy pedestrian area may require examination of crosswalk conditions, traffic signals, nearby surveillance footage, witness accounts, vehicle sensor data, object-detection records, braking commands, and the autonomous vehicle’s operational limitations.
The local setting matters, but California’s statewide negligence, products-liability, insurance, and autonomous-vehicle rules remain central to the legal analysis.
Steps to Take After a Driverless Vehicle Collision
After a serious crash involving an autonomous vehicle, practical steps may include:
- Seek appropriate medical attention.
- Report the collision to law enforcement when required or appropriate.
- Photograph the autonomous vehicle, license plate, vehicle identification information, damage, road, traffic signals, and surrounding conditions when it is safe to do so.
- Identify the autonomous-vehicle company or fleet operator.
- Preserve witness contact information.
- Save any ride-app records, receipts, trip confirmations, messages, and screenshots.
- Preserve your own dashcam or mobile-phone recordings.
- Obtain the collision report.
- Avoid altering or disposing of damaged property that may become evidence.
- Consider early preservation requests for vehicle and electronic data in a serious injury case.
Our broader California accident claim guide explains how liability, evidence, insurance, medical documentation, and damages fit together after a serious collision.
Deadlines Still Apply to Autonomous-Vehicle Injury Claims
Autonomous technology does not eliminate California filing deadlines. For many personal injury actions based on wrongful conduct or negligence, California Code of Civil Procedure section 335.1 provides a two-year limitations period.
That rule should not be treated as universal. Different deadlines may apply depending on the defendants, legal theories, age or status of the injured person, government involvement, contractual issues, or other circumstances. Claims involving a public entity can also trigger separate government-claim requirements and substantially shorter procedural deadlines.
In a technology-heavy case, waiting can also create a separate practical problem: valuable evidence may become harder to locate even before the legal filing deadline expires.
Frequently Asked Questions About Driverless Car Liability
Is the passenger in a driverless taxi responsible for a crash?
Usually, merely being a passenger does not make someone responsible for the autonomous vehicle’s driving decisions. Liability would depend on the passenger’s actual conduct and the circumstances. The investigation instead commonly focuses on the automated-driving system, vehicle or component manufacturers, fleet operator, and other road users.
Can an autonomous vehicle manufacturer be sued after a crash?
Potentially. A manufacturer may face a claim when evidence supports an applicable theory such as defective design, manufacturing defect, failure to warn, or negligence. The presence of autonomous technology alone does not establish a defect or liability.
Can both the technology company and another driver be liable?
Yes, depending on the evidence. California’s comparative-fault system allows responsibility to be allocated among multiple parties when more than one person’s or entity’s conduct contributes to an injury.
Does California require autonomous vehicles to record crash data?
California Vehicle Code section 38750 requires qualifying autonomous vehicles to contain a mechanism that captures and stores specified autonomous-technology sensor data for at least 30 seconds before certain collisions and preserves that data for three years. Additional electronic information may also exist outside that statutory dataset.
Is Tesla Autopilot legally the same as a fully driverless robotaxi?
Not necessarily. California law distinguishes Level 3 through Level 5 autonomous vehicles from systems that provide driver assistance while still requiring active human monitoring or control. The specific vehicle, software version, automation level, and system status at the time of the collision must be identified rather than relying on marketing terminology alone.
Can a driverless-car company refuse to provide its electronic data?
An injured person may not automatically receive all proprietary vehicle or software information simply by requesting it. Depending on the circumstances, obtaining evidence may require preservation notices, formal discovery, subpoenas, court orders, inspections, or litigation. Identifying which entity controls each category of data is therefore an important early step.
A Driverless Car Case Is Ultimately an Evidence Case
The absence of a human driver does not mean the absence of potential responsibility. It changes where the investigation needs to look.
A serious autonomous-vehicle case may require reconstructing what the vehicle saw, what its software predicted, whether its sensors functioned correctly, how it responded, whether the system remained within its operating limits, whether another road user contributed to the crash, and which companies controlled the relevant information.
For that reason, determining liability may require coordination among accident reconstruction professionals, automotive engineers, software or electronic-data professionals, medical providers, insurance professionals, and attorneys.
Speak With The Sterling Firm About a California Driverless Car Crash
The Sterling Firm represents people injured in California motor-vehicle accidents, including cases involving complicated liability, multiple responsible parties, insurance disputes, defective products, and electronic evidence.
A California personal injury lawyer can investigate the circumstances of a serious autonomous-vehicle collision, identify potentially responsible parties, preserve available evidence, examine insurance coverage, and determine which legal theories the evidence may support.
FREE PERSONAL INJURY CASE EVALUATION
NO UPFRONT ATTORNEY FEES
NO FEE UNLESS WE WIN
Call or text The Sterling Firm at (310) 498-2750 or toll free at (844) 4-GETLEGAL (844-443-8534). Attorney Advertising. This article provides general educational information and is not legal advice. Results vary, and reading or contacting the firm through this article does not by itself create an attorney-client relationship.
Pasted text
Do you like this personality?






