Skip navigation
Insuring Self-Driving Cars

Insuring Self-Driving Cars

There will be a strong incentive to find that fault in an accident rests in an highly automated vehicle, not the human driver, especially if there are serious injuries.

Self-driving cars are expected to have fewer and less-severe accidents.

However, accidents still will happen, and as cars become increasingly automated, insurance for traffic accidents will change.

Currently, the cost of most traffic accidents is paid by auto insurance. Private passenger coverage is purchased by individuals; commercial auto coverage is purchased by businesses and typically covers fleets of vehicles.

In virtually every state, a minimum level of liability coverage is required by law to cover damage an insured driver may cause to other people and vehicles. Purchase of first-party auto coverages to protect the insured’s own vehicle from damage or theft loss (collision and comprehensive) are optional, as are higher limits of liability coverage.

Occasionally, a series of accidents will be attributed to a vehicle defect and lawsuits will be filed against the manufacturer, potentially triggering a claim under the manufacturer’s product-liability insurance. If the product is found to be defective, the auto insurers that paid claims for accidents caused by the defect will bring indemnity claims (termed “subrogation”) against the manufacturer’s product liability insurer.

If the manufacturer is self-insured, which is typical for very large enterprises, the subrogation claim for indemnification will be brought against the manufacturer. Self-driving cars will transform this current pattern in which most traffic accident costs ultimately are paid by auto insurers.

The costs of accidents involving autonomous cars that retain driver controls, and which require the driver to retake control if the automated systems cannot cope with the driving conditions, will be split between auto insurers and product-liability insurers (or self-insured manufacturers).

When an autonomous car with driver controls collides with another vehicle, the question will be: Was the autonomous vehicle the cause of the accident? If the answer is yes, a series of questions will follow: Was the driver or the software in control? Did the driver ignore a signal to take back control of the car? Was the driver given enough time to reasonably respond to the alert?

There will be a strong incentive by all participants in the accident to find that fault rests in the autonomous vehicle, not the human driver, especially if there are serious injuries. The product-liability policy (or self-insuring manufacturer) will be a much deeper pocket than the auto policy, and the human driver will avoid being assessed driving record points that would increase his or her insurance costs.

Insuring Fully Autonomous Cars

Accidents involving fully autonomous vehicles that do not have steering wheels, accelerator pedals or brake pedals predominantly will generate product-liability claims. Auto liability insurance likely still will be required under law for fully autonomous cars and a handful of claims still may trigger only the auto liability policy – but the facts would be quirky.

For example: Your teenagers decide to push your self-driving car across your sloping driveway. The car rolls down the driveway into the street and collides with another car. The self-driving car was in “park” and its parking brake was deployed; thus, there was no basis for a product-liability claim.

People will continue to purchase first-party auto coverage. They will want to recover for the damage to their expensive self-driving car if a tree branch falls on it or another vehicle collides with it.

When a fully autonomous car collides with another vehicle, the first question again will be: Was the autonomous vehicle the cause of the accident? If the answer is yes, a different series of questions will follow: Which system failed? Was it a sensor problem? A software problem? A data download problem? A hacking issue? A combination of problems?

The answers to these questions will be of critical interest to claimants and insurers, as well as engineers. It is likely that auto manufacturers will obtain sensors, software and data updates from other entities. Each of these entities will have its own product-liability policy or be self-insured.

If the accident was caused by hacking, cybersecurity policies of the manufacturer and/or a supplier will be triggered.

Determining which system failed and, therefore, which company’s insurance should pay, will rely to a great extent on the “black boxes” that will be required on all autonomous vehicles. Whether the black-box data will be sufficient to reconstruct accidents at the granular level needed to determine the precise nature of the failure is unknown.

As the percentage of accident costs paid by auto insurance decreases and the percentage paid by product liability insurance increases, the cost of accidents will shift to the purchase or lease price of the car. Self-driving cars not only will reduce total accident costs, but also will change how we pay for them.

Hilary Rowen is a partner in the San Francisco office of Sedgwick LLP. Her practice focuses on insurance regulatory issues. You can reach her at [email protected].

TAGS: Vehicles
Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.