Insurers weigh who to blame when a self-driving car crashes
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Brendan Lynch/Axios
As cars begin to drive themselves, the onus of liability gets murkier, which is creating new headaches for the insurance industry.
Why it matters: More than half of all new cars in the U.S. will offer hands-free driving by 2028, according to Telemetry's 2025 Assisted and Automated Driving Forecast.
- For insurers, that means rethinking risk assessments based on whether humans or robots are driving and writing policies that separate driver liability from that of the vehicle itself.
The big picture: Crash avoidance systems like automatic emergency braking, lane-departure warning and blind-spot detection have proven safety benefits, according to the Insurance Institute for Highway Safety.
- That could, in theory, mean fewer insurance claims.
- But drivers sometimes turn them off, or use technology incorrectly, and so crashes still occur.
- When they do, these advanced vehicles typically cost more to repair because of all the expensive sensors, cameras and chips that enable those technologies.
By the numbers: The average repair cost exceeds $4,700, up more than 96% since 2009, writes Kyle Krumlauf, director of industry analytics at CCC Intelligent Solutions.
- Higher repair costs are also due to the rise of electric vehicles, which are heavier and cause more damage in crashes. Plus, EV replacement parts can be expensive.
Between the lines: Insurers need to understand the risks associated with these modern vehicles in order to set rates on their policies.
- But that's where it gets tricky.
- "The whole heart of insurance is being able to evaluate risk and figure out what they're insuring," says Brett Odom, policy vice president for auto and alternative vehicles for the National Association of Mutual Insurance Companies.
How it works: Current underwriting practices rely on driver behavior, accident history and traffic offenses.
- These insurance models are rooted in actuarial studies that link prior behavior to the likelihood of future accidents.
- That data is then used to help formulate risk and, ultimately, to set insurance rates.
The intrigue: As driving control shifts to AI systems, insurers must also analyze vehicle behavior.
- But software systems change frequently via over-the-air updates.
- Even more challenging are scenarios where drivers and vehicles pass control back and forth.
- While today's hands-free highway-driving technology requires drivers to pay attention, future systems will let them take their eyes off the road under certain conditions.
- If the driver isn't required to be in the loop, liability will likely shift to the vehicle manufacturer or software provider — representing a fundamental change in the structure of auto insurance policies.
What to watch: One of the biggest hurdles to insuring automated vehicles, however, is getting access to vehicle data, according to a report by S&P Global Mobility.
- Without regulatory mandates or industry-wide standards, carmakers retain control over vehicle logs and system data.
- That makes it hard for insurers to investigate claims and set prices.
- "If you're the owner of the vehicle, you should have the ability to decide who has access to that data," Odom said.
Where it stands: For insurance companies, it's still early to predict what impact driverless cars will have on the insurance industry, notes Progressive.
- For now, vehicles with automated systems are still subject to the same state minimum car insurance requirements as other vehicles.
