Function-specific Automation (Level 1): This category covers a wide range of features in common use, including cruise control and antilock brakes.

Combined Function Automation (Level 2): At least two primary control functions — such as steering and acceleration or distance maintenance and lane maintenance — are automated and designed to work in unison. With the more sophisticated Level 2 cars — which are not yet available to the public — the driver can cede active primary control in limited situations but must remain constantly attentive and ready to take control at all times and on short notice.

Limited Self-Driving Automation (Level 3): The vehicle has sufficient self-driving capabilities to enable the driver to fully cede control of safety-critical functions, in many driving conditions. The vehicle signals when the human driver needs to reengage. The driver retains the ability to take control of the car.

Full Self-Driving Automation (Level 4): The car — perhaps better described as a “street legal robot” — can drive itself in all conditions, does not require a driver and may not have any driver controls (other than a destination entry pad and a “go” button).

Auto Accident Litigation and Insurance Coverage Disputes When Cars Drive Themselves

Despite the potential of self-driving cars to significantly reduce auto accidents and injuries, autonomous and semi-autonomous vehicles do not fit neatly into the current motor vehicle and insurance regulatory regimes. In response to the emerging issues posed by self-driving cars, the California Department of Motor Vehicles and the Department of Insurance are responding. The DMV has issued regulations, effective September 16, 2014, governing the testing of autonomous vehicles on public roads[1] and has indicated that regulations governing autonomous vehicles offered for sale to the public will likely be issued by the end of 2014. Insurance Commissioner Dave Jones held a public hearing September 15 on insurance issues posed by autonomous and semi-autonomous vehicles. The hearing was streamed live and can be viewed at http://www.insurance.ca.gov/.

Today, most of us drive a Level 0 or Level 1 car. Cars that meet the bare minimum of the Level 2 criteria are starting to enter the market. A fairly rapid acceleration through Level 2 is likely during the next five years, as more functions become automated and more car makes and models include multiple interacting automated functions.

When drivers can legally start “multitasking” while behind the wheel, the transition from Level 2 vehicles to Level 3 vehicles will occur. “Hands-free” driving, in which the driver can text, talk on a phone or read a book, will be lawful, at least in some circumstances. However, the driver will still be required to take back control when signaled by the vehicle. At Level 4, the street legal robot is programmed to go to its destination without a driver and, if desired, without a passenger. Your street legal robot could take you to the airport while you check your email and then drive itself home.

Novel Insurance Coverage Issues

Autonomous vehicles will be safer than cars under driver control, and they may reduce the occurrence of accidents. Nevertheless, self-driving cars – especially Level 2 and Level 3 cars – are likely to pose novel insurance coverage issues. Consider the following accident scenarios:

A: A vehicle operating in autonomous mode is involved in a two-car accident. The vehicle in autonomous mode is alleged to be at fault because it ran a light that was turning red. There is a fact dispute concerning whether a sensor or the software technology failed or the driver took control from the automated functions in order to run the light.

Here, both of the parties in the accident may want to assert that the cause was a technology failure. The driver of the semi-autonomous car may want to avoid an “at fault” finding that would generate “points” – at least under the current auto insurance rating regulations. The other driver (and any injured passengers or pedestrians) also may want to contend that the accident was caused by a technology failure. Any claim against the at-fault car would be capped, as a practical matter, by the policy limits of the at-fault car's auto policy. In contrast, a product liability claim against the auto manufacturer could seek full economic damages.

Data retained in the car's “black box” could resolve the determination of causation. Black boxes – more formally known as event data recorders (“EDRs”) – are common in recent model-year cars. EDRs typically are programmed to record data in a continuous loop, writing over information again and again until a vehicle is in a collision in which the airbags deploy. Vehicle Code Section 38750(c)(1)(G) mandates that Level 3 and Level 4 vehicles have a “black box” that retains at least 30 seconds of data immediately prior to a collision. As driving becomes increasingly automated, “black box” data will become more crucial to the resolution of auto claims.

B: A car is in autonomous or semi-autonomous mode: the driver is texting. The car signals the driver to retake control due to adverse weather or a sudden change in road conditions. The driver is disoriented and slow to respond, and a one-car accident occurs. The driver brings a product liability claim, asserting that the car did not provide adequate notice of the need to retake control.

If the car is a Level 3 autonomous vehicle, the driver is allowed to engage in other activities while behind the wheel. The product liability claim may well be viable. The outcome may depend not only on the length on period between the alert and the point at which the autonomous function shut down, but also whether the driver had reason to anticipate that driver control would be required (for example, entering snowy conditions or very heavy rain that can limit sensor range.

Although the product liability claim would likely fail if the accident involved a Level 2 semi-autonomous vehicle with the driver responsible for remaining alert and being capable of taking control, drivers are likely to treat a car at the high end of Level 2 as capable of fully autonomous operation. Sophisticated Level 2 cares will be able to maintain distance in traffic – braking and accelerating as needed – and can keep within a lane without the driver's hands on the wheel. The temptation to send a quick text is likely to be irresistible, at least to some drivers.

Basic Liability Rules Do Not Need to Change

Novel accident causation theories and coverage disputes generated by vehicle automation will not change the questions raised by basic tort issues: What happened? Who was at fault? How much was each party at fault? At the core level of legal analysis, the brave new world of self-driving cars will be familiar territory for lawyers. However, the more mundane aspects of practicing law will change in the not too distant future. Imagine being able to read expert statements or deposition transcripts while en route to the courthouse – or being able to review a draft contract en route to a client meeting – while your car drives itself.

Note

About the Author

Hilary Rowen is a partner in the Insurance Regulatory Practice at Sedgwick LLP's San Francisco office. She was one of nine experts invited to testify at the California Department of Insurance's public hearing on autonomous vehicle insurance issues.

Ms. Rowen may be contacted at [email protected].