Driverless vehicles are currently being developed and tested on American roads, and they’re making lots of headlines along the way about both their promise and their problems. Your current car may even have some of the same features that will help guide future self-driving cars. Proponents note that autonomous cars may lead to safer roadways by eliminating the most common cause of accidents: driver error.
But autonomous vehicles also present challenges regarding safety, risk, and insurance. Discover what’s going on now with driverless cars—and what the future may hold.
The Market for Driverless Cars
Right now, engineers are at work creating driverless experiences—and driverless cars. In fact, the car or SUV you drive today might include features that will be integral to autonomous vehicles in the future. In fact, testing and deployment are ongoing in many states, including Arizona, Florida, New York, Ohio, and California. But self-driving vehicles are not yet available for consumer purchase.
Levels of Automation
There are five automation levels with varying car features, according to the Society of Automotive Engineers and the National Highway Traffic Safety Administration (NHTSA).
- Level 1 mostly depends on the driver but may offer warnings or some automated assistance, such as adaptive cruise control.
- Level 2 is partially automated, with a combination of advanced driver-assistance systems (ADASs) controlling the car’s speed and braking.
- Levels 3 to 5 increase the car’s amount of control over braking, steering, and route selection. While autonomous vehicles at level 3 and above are being developed by various companies, the NHTSA notes that we are “many years” from publicly accessible, fully automated cars.
“Currently, we only have level 2 or partially autonomous vehicles available commercially,” said Marc Gordan in an email to The Balance. Gordan is the senior director of the Connected Car team at LexisNexis Risk Solutions, which provides data and analytics solutions to the insurance industry. These ADASs can navigate highways and traffic and perform basic tasks, such as changing lanes and entering and exiting parking spaces, he noted.
But the systems are still evolving. “The main difficulty for the roads of tomorrow is that a fully autonomous vehicle must handle those elements in addition to busy city streets and poorly lit and twisting country roads throughout the range of seasons—all without fail,” Gordan explained.
The Road to Progress
“Introduction of autonomous vehicles is expected to be slow,” said Karlyn D. Stanley, who studies autonomous vehicles as a senior policy analyst at the RAND Corporation, to The Balance. “The lifespan of a conventional vehicle is 11 years, so the composition of cars on the road is going to be a slow transition. It’s not going to happen overnight.”
As part of her work at the nonprofit think tank, Stanley spoke with 43 experts across the insurance and autonomous vehicle industries for a December 2020 RAND report, “Autonomous Vehicles and the Future of Auto Insurance.” Most felt the initial introduction of autonomous vehicles would be in the fleet model, such as found in ride-sharing or delivery services, noting that while autonomous cars will likely cost more at first, the expense is balanced by the scale of a fleet, which may also cost less to insure due to the scale.
The Risks of Self-Driving Vehicles
Various companies are developing autonomous driving technology and testing vehicles, including WeRide, Argo, Zoox, Lyft, Cruise, Waymo, and others. Many of these companies partner with or have been acquired by well-known automakers, including Honda, Ford, and General Motors. But self-driving vehicles won’t be overtaking country lanes and highways anytime soon due to hurdles and challenges. Various risks need to be considered by legislators, drivers, and insurers.
In June 2021, NHTSA began requiring manufacturers and operators of vehicles equipped with ADASs or levels 3-5 automated driving systems to report crashes promptly. In California, any collisions that cause injury, death, or damage must be reported to the state’s Department of Motor Vehicles, and as of August 2021, the state had fielded more than 300 reports. Automakers are attempting to remedy safety concerns, as well as to understand how human responses influence safety.
Despite the low number of daily driver-error accidents, autonomous-vehicle accidents make headlines. In 2018, a pedestrian in Arizona was killed by a test vehicle operated by Uber Technologies that had a human operator at the helm. Although the National Transportation Safety Board report alleged the operator’s failure was the probable cause, it also stated the vehicle didn’t classify the pedestrian or predict the pedestrian’s path.
Self-driving cars will generate, collect, and retain large amounts of personal data about their users. Many cars will communicate with one another and with roadway infrastructure while on the road. As a result, there are valid and yet-unsolved concerns around who owns and can view consumer data. The car will also rely on technology for safe operation and be potentially vulnerable to hacking through its electronic system.
Insurance and Self-Driving Cars
In an evolving industry, the insurance landscape around autonomous vehicles will continue to develop as well. Better understanding how automation interacts with the driver will help determine who’s liable in an accident: the human driver, the machine/software, or the automaker. Right now, a patchwork of federal and state guidelines and legislation governs the testing and operation of autonomous vehicles, along with their insurance requirements.
“In the future, automakers will likely absorb more liability, but it will be interesting to observe how [the] year, make, and model fit into the overall picture,” Gordan said. “Similar to other industries impacted by innovation, insurance carriers will look to protect their revenue in changing times.”
Liability might shift from the consumer to the automaker, for example, and insurance carriers might create new products for consumer purchase, he said. “As the vehicle absorbs more driving responsibility or more variables enter the equation, like smart streetlights, there are bound to be changes to how the ecosystem manages risk.”
The NHTSA offers federal guidelines for automated driving systems, while states are in various phases of regulating testing, public road operation, vehicle data privacy, and insurance.
Here are some of the pending insurance concerns:
Insurance Frameworks and Liability
Most experts Stanley spoke with thought that today’s auto insurance would adapt to driverless cars and there wouldn’t be significant changes in the near future. For businesses, the fleet insurance model seemed most practical and achievable. With this framework, an automated car manufacturer might sell or lease a group of self-driving cars to a delivery service, which would then either self-insure or add the fleet to a corporate policy to pay for liability claims. Repairs would be performed in-house, another way to minimize costs.
Currently, insurance is regulated at the state level, and many states are increasingly wading into insurance and liability concerns regarding self-driving cars, although some states have no legislation.
In Washington state, laws require the testing company to have a minimum $5 million umbrella liability policy to cover any accidental injury, death, or property damage. In Oklahoma, liability coverage isn’t addressed by state law, while no liability coverage is required in Virginia.
While accident rates may decrease, repairing a self-driving vehicle after a minor fender bender could add up for insurers, the RAND report explained. Autonomous cars rely on many sensors and cameras and will likely require specialized auto shops that can repair and calibrate this equipment.
“This isn’t something you can handle in your garage,” Stanley said. Fleet operators—such as businesses with taxis or rental vehicles—already use in-house maintenance and can more easily deal with repairs. The manufacturers may also centralize repairs.
Most experts feel there will be more demand for cybersecurity insurance, according to Stanley’s report. However, the insurance industry thought they could meet demand and set accurate rates. Future annual policies might include or allow you to add a cybersecurity rider to your annual policy for an autonomous vehicle.
Claims and Safety
Insurance companies are still trying to weigh the benefits and risks of existing automated processes to help set accurate policy rates. Whether claims will go up and down also depends on how automation influences safety. Insurers are often looking to ADASs for clues.
“We found that ADAS-equipped vehicles show a 27% reduction in bodily injury claim frequency and a 19% reduction in property damage frequency,” Gordan said. The coexistence and combination of features—more common on new models—can influence outcomes, he said.
Gordan added that consumers with these features expect their auto choice will positively impact their insurance premium. “The benefit for consumers is great because we've seen reductions in claims and, on average, discounts of up to 10% off their insurance premiums on average,” he said.
Insurers may potentially consider driving behavior data and ADAS variables when underwriting a policy, such as whether the ADAS is on or off, how it’s configured, whether the driver is wearing a seatbelt, or whether the headlights are on.
The Highway Data Loss Institute also found that technological improvements reduced insurance claims regarding property damage and bodily injury, usually when a combination of features are engaged. This is primarily the case in second-level automation, such as forward collision warning, lane departure warning, and automatic emergency braking.
But there may be a few twists in the road en route to a safer, less expensive outcome. A recent study from IIHS found that drivers using adaptive cruise control in conjunction with lane-centering technology often set the speed above the speed limit and decreased the amount of attention they paid to driving—thereby giving up safety benefits. Other automated safety features can be deactivated when confronted with curves in the road.
The Future of Driverless Car Insurance
“Driverless cars are in our future, there's no question about it,” Gordan said. “How and when will be the big questions to consider as intricate planning and collaboration are required to build smart city infrastructure, artificial intelligence advancements, regulatory impacts and, of course, consumer acceptance.”
According to one study, 35% of survey respondents said they’d be more likely to buy an autonomous vehicle if doing so led to decreased insurance rates.
Frequently Asked Questions (FAQs)
How do self-driving cars work?
Autonomous cars rely on various sensors, computers, and cameras to guide them through the driving environment to a destination. For example, GPS sensors could direct a vehicle to the nearest gas station, while a camera and sensors would detect objects, predict their behavior, and respond to them—whether those objects are pedestrians crossing the road or the fuel station’s pumps.
When will self-driving cars be common?
Self-driving cars won’t be available anytime soon, according to the NHTSA. Manufacturers are addressing tech- and cybersecurity-related issues, and rigorous testing must occur before autonomous vehicles are placed on the showroom lot. However, some future features of autonomous vehicles are available now, such as sensors that detect and act, like emergency braking, lane assist, and adaptive cruise control.
Why do we need self-driving cars?
Autonomous cars offer many potential benefits, according to the NHTSA, including:
- Increased safety due to reduced human error, the primary cause of auto accidents
- More efficiency and convenience due to reduced congestion.
- Increased mobility for older people and people with disabilities that prevent them from operating a motor vehicle.