Auto Insurance for Self-Driving Cars

Autonomous technology will have big consequences for both the insurance industry as a whole and individual plans, but its full impact remains to be seen.

Modernizing the Regulatory Environment

The regulatory environment is still catching up with autonomous technology and insurance. Various levels of government, from federal to local, play a role in regulating self-driving cars. Thus far, the federal Department of Transportation (DoT) has not issued comprehensive rules for insuring an automated vehicle, but rather has imposed voluntary safety guidelines for car manufacturers.¹ It’s possible that the DoT will release guidance for insurance providers as well, but that remains to be seen.

With this federal inaction, it has largely been up to individual states and municipalities to impose rules about insuring self-driving cars. According to the National Conference of State Legislatures, 29 states have enacted legislation related to autonomous vehicles so far.² Insurance requirements vary widely by state: Some have no legislation, while others require consumers and testing companies to have umbrella liability.

For the most part, drivers whose cars utilize automation simply need to follow their state’s auto insurance requirements. This will change as regulations evolve and technology progresses, however.

Shifts in Underwriting

Underwriting is the way insurance companies evaluate their risk in offering coverage. Traditionally, providers take various factors into account, from an individual’s driving history to their location. Advances in self-driving vehicles may lead underwriters to focus more on the make and model of the car, as the technology becomes more important than the individual sitting in the driver’s seat. In the future, driverless car manufacturers may release a set of risk factors, based on extensive real-world testing, that provide underwriting criteria across all markets.

Redefining Fault: Claims and Liability

After an accident, insurance companies determine the actor at fault, which establishes liability. Self-driving cars turn this formula on its head. Some of the most interesting questions around autonomous technology center on liability.

Whose “fault” is it when there is an incident involving a self-driving vehicle? Current laws typically punish human drivers, but what about manufacturers, software developers, and sensor vendors? Are they to blame if a fully autonomous vehicle crashes, rather than the human passenger?

Insurance providers will need to answer these questions as self-driving technology progresses. Today’s autonomous vehicles still require human attention and intervention, meaning a person can still be held liable in the event of a malfunction that leads to an accident. But experts predict that vehicle and software manufacturers will be held liable in at-fault collisions in the future.

Time will tell how this space develops. Though autonomous safety features like collision warnings and emergency brakes promise to reduce accidents, liability coverage will still be necessary for when something goes wrong.

The Future of Car Accidents

What if you get into an accident with an autonomous vehicle? For now, treat it like a normal car accident — don’t leave the scene, document everything, exchange information with the other parties, and report it to your insurance company.

We can only speculate about how car accidents will look in the future. They may occur far less frequently thanks to removal of human error. When they occur, it’s possible that all the relevant information will be reported directly to both the vehicle manufacturer and the insurance company, simplifying the process for everybody involved.

Automation and Insurance Premiums

It’s still not clear how autonomous vehicles will impact insurance rates and premiums. Some predict that rates will go up, since the technology that enables automation (sensors, cameras, computers) is expensive to source and repair. But self-driving cars are also expected to lower accident rates, meaning there is less to fix.

These two factors (cost to repair and increased safety) may offset each other, causing rates to remain the same. The price of insuring a self-driving car is ultimately an open question. Today, your auto insurance cost will be the result of all the standard factors, such as mileage, location, and driving history.

Current Insurance Options for Self-Driving Cars

Even though we are still waiting for full automation and resulting guidance, most major insurance providers already offer coverage for cars with some level of driving assistance. Here are our reviews of a few companies that cover vehicles with assistive driving features:

• Allstate review
• Nationwide review
• Liberty Mutual review
• Travelers review
• Esurance review

Tesla also offers its own insurance product. The company’s pricing structure takes into account automated assistance features as well as individual driving habits, which could improve safety and lower rates. Though it’s only available in California and Texas, this plan may be the best auto insurance for Tesla.

Driving automation technology is extraordinarily complicated and still presents a challenge to researchers. At the base level, though, self-driving cars work by using an array of sensors to model their surroundings and make split-second decisions about how to respond.

Sensors typically include lidar, GPS, lasers, and camera arrays that cover multiple surfaces of the car. Software then processes these inputs and constructs a 360-degree “view” of the environment around the vehicle.

Next, the computer plots a safe path and directs the car to follow it. This occurs multiple times per millisecond. The software must be impeccable at predicting movement and function in diverse conditions.

For nearly as long as motorized vehicles have existed, researchers have dreamed of a car that could drive itself. As far back as 1925, an inventor drove a radio-controlled car through the streets of Manhattan without anybody at the steering wheel.

The technology behind modern driving automation, neural networks, was first theorized in the early 1990s. Various companies tried to develop the new technology, but Google is typically credited as a leader in self-driving technology with the formation of Waymo in the 2000s. After that, major car manufacturers jumped into the fray, and today the race is on to see which company will break through with truly autonomous technology.

There are various definitions of autonomous technology that impact what we mean when we talk about “self-driving” cars. Not every autonomous feature is available today.

Levels of Automation

The National Highway Traffic Safety Administration has defined six levels of automation based on a car’s ability to make decisions and act independently.

• Level 0: The driver performs all tasks with no automation.
• Level 1: The most basic level of automation, this tier includes warnings and partial driving assistance (think adaptive cruise control).
• Level 2: When a car can control speed and steering on its own under certain circumstances, it falls into this category. This includes advanced driver-assistance systems (ADAS) like lane assist and automatic parallel parking.
• Level 3: Conditional automation means that the car can perform tasks like steering, braking, and acceleration in certain situations, but a driver must be ready to take control at all times.
• Level 4: The car can drive itself with no human intervention in certain conditions.
• Level 5: This is full self-driving, meaning the vehicle can drive itself with no human required at all.³

Autonomous Features Available Now

Level 1 features are now widely available in consumer cars, including adaptive cruise control and land centering. Level 2 is the highest level of automation you can buy commercially. This includes well-known technology like Tesla Autopilot and GM Super Cruise that can control a car under certain conditions.

Most manufacturers are currently developing their own versions of this consumer technology. However, it remains expensive and incomplete, as a driver must be ready to take over at any time. Full automation (Level 3 and above) is not yet available for consumer purchase.

Full driving automation has the potential to bring many benefits to consumers. Here are some examples.

• Safety improvements: A computer cannot be distracted by a cell phone or drive under the influence. Besides removing human error from the equation, safety features like collision warnings and emergency braking could reduce accidents, making drivers and pedestrians safer.
• Fuel efficiency: Self-driving cars can plot the most efficient routes and modulate their speeds based on factors like predicted wait times at stoplights. This means less gas will be wasted idling.
• Traffic reduction: A fleet of self-driving cars could communicate with each other and reduce congestion on roadways.
• Time savings: A self-driving car could not only be faster and more efficient, but also free up human time to focus on other things. Imagine catching up on emails while your car drives you to the office.

Currently, researchers are working on the technology to enable levels 3, 4, and 5 of automation. It’s not clear how much longer we have until full automation becomes available, let alone widespread.

When Will Full Automation Become Available?

Business leaders like Elon Musk have been predicting and even marketing Level 3 and higher cars for several years, claiming that the technology is just around the corner. A former U.S. secretary of transportation even said we’d have widespread autonomous cars by 2021,⁴ clearly an underestimate.

But some experts, including many in the insurance industry, think that full automation is still years away, as many practical and technical problems need to be solved first. Manufacturers will have to contend with these issues before truly autonomous cars can be accepted.

What’s holding the car industry back from full driving automation?

Safety

This is a crucial consideration. Recognizing and distinguishing between objects remains a difficult task for computers. Researchers are continuing to improve the artificial intelligence that makes such decisions, but it still needs to be verified before it hits the road for consumers. After all, distinguishing between a cyclist and a parked car is a life-and-death decision for everybody involved.

Fatalities Involving Self-Driving Cars

The first known fatality involving a self-driving car occurred in 2016 with a Tesla in Florida. Since then, cars operated by computers have caused several deaths. Fatalities, and the difficulty determining fault for them, underline the moral challenges that accompany autonomous driving technology.

Hacking Risk

No computer is impervious to malicious cyberattacks. Given their reliance on computers, self-driving cars are more vulnerable to hacking than other vehicles.

This presents some frightening scenarios. What happens if malicious actors decide to shut off a car’s brakes, for example? Self-driving cars must include robust cybersecurity.

Public Perception

Consumers will have to buy into any shift in technology. Public perception of self-driving cars is mixed: Some drivers can’t wait to have their car drive them everywhere, while others fear giving up control. This may be a limiting factor in adoption of fully autonomous vehicles.

If you’re confused about the state of self-driving cars, you’re not alone. The technology is complex, and so are the questions it raises. But for now, if you buy a car with autonomous features, you can rest assured that you will find an insurance plan to cover your needs. As the technology develops, so will the industry, and you will adapt alongside everybody else.