Insurance for AVs will rely on sophisticated AI

Illustration: Lazaro Gamio/Axios

The insurance industry is learning how to incorporate artificial intelligence in various scenarios to determine liability and calculate risk, including AVs, where the responsibility for malfunctions and accidents could fall on safety drivers, on the vehicle manufacturer, or on software and hardware providers.

Why it matters: Insuring AVs is an unprecedented insurance challenge, as they will generate huge volumes of data from a variety of parts and will be comprised of complex systems that share responsibility in interlocking ways.

What's needed: Regardless of what’s behind a claim, insurance relies on past data for predictive analytics.

  • Because AVs do not have decades of safety records, insurers will need programs that can extrapolate as much insight as possible from a small number of initial incidents.
  • Once initial datasets are established, insurers will still need AI that can sift through all of a vehicle's relevant data from each incident reported to assign responsibility.

What's happening: AI is already used in catastrophic storm insurance, another insurance challenge that involves unprecedented events, and sifting through massive amounts of data. AI-based systems might work for AVs in multiple ways:

  • Insurers could deploy advanced computing from the moment a customer files a claim to begin sorting through vehicle data from the incident in question.
  • AI could prioritize and investigate the most important questions at a given point in time, and combine reported claims with weather reports and data from on-the-ground sensors to get the full context of an incident.
  • Using AI throughout the claims review process could help risk managers spot patterns and anticipate other reports that may follow, whether that's because of a major weather or traffic event or because of a faulty component within an AV's tech stack.

The bottom line: Insuring AVs will rely on numerous AI-based tools for data analysis: for building out a preliminary database, for analyzing individual incidents, and for identifying patterns in order to improve safety going forward.

Haywood Marsh is general manager of NetClaim, an insurance software subsidiary of NAVEX Global.

What's next

⚖️ Live updates: Democrats' second day of opening arguments

"He could release the aid or break the law. He chose to break the law," House impeachment manager Rep. Zoe Lofgren (D-Calif.) said at day 3 of Trump's Senate impeachment trial.

The big picture: Democrats have roughly 8 hours remaining over the next two days to lay out their case against Trump's alleged abuse of power and obstruction of Congress.

Go deeperArrowJan 21, 2020 - Politics

Putin foe Bill Browder was warned of potential danger in Davos

Photo: Julia Reinhart/Getty Images

No story caused a bigger stir in Davos this week than the news that two suspected Russian spies had been caught in August posing as plumbers in the Alpine town.

Between the lines: One prominent attendee instantly suspected a personal connection. Bill Browder, a U.S.-born financier and long-standing thorn in the side of Vladimir Putin, tells Axios that before departing for Davos he received a warning from the British security services — passed along by their Swiss counterparts — that he could be in danger.

Go deeperArrow5 hours ago - World

Chinese cities cancel Lunar New Year events amid Wuhan coronavirus outbreak

People in Guangzhou, China, wearing masks on public transportation. Photo: Stringer/Anadolu Agency via Getty Images

25 people have died from an outbreak of a new coronavirus strain that originated in Wuhan, China, including the first outside the immediate area, provincial authorities said Thursday.

The latest: Major cities in China, including Beijing, have canceled large public gatherings for the Lunar New Year holiday, the most important in the country, to help contain the outbreak, according to the Washington Post.

Go deeperArrowJan 20, 2020 - World