Geico to Use Artificial Intelligence to Speed Up Car Repairs

Financial terms of the partnership weren’t disclosed.

Todd Combs,

Geico’s chief executive, said in a written statement that Tractable’s technology is a way to obtain accurate estimates for policyholders and get drivers back on the road faster.

Geico is second in the private-passenger auto-insurance market, with a 13.6% share, according to the Insurance Information Institute. Geico’s size as a part of

Warren Buffett’s

Berkshire conglomerate means its moves are often followed by other car insurers, so its use of artificial intelligence in handling claims could become standard industry practice.

A spokeswoman for market leader State Farm Mutual Automobile Insurance Co., with a 16.2% share, said the insurer uses some artificial intelligence as part of its claims process for damaged vehicles. “When a customer files a claim, machine learning predicts with a level of confidence if a vehicle is a total loss or repairable,” she said. Human claims adjusters are still involved, she added.

Under Tractable’s approach, an auto-body shop would send in photographs of a damaged vehicle to Geico with its repair estimate. From there, algorithms would assess the severity of the damage and any necessary repairs, which the algorithms have learned from reviewing millions of historical estimates and photos, Tractable said. The repair estimate could be confirmed within minutes.

If the artificial intelligence notices an issue with the estimate, it provides information to a human Geico adjuster, who can then review the damage and repairs with the shop. Repairs can start immediately if the estimate is confirmed, quickening the process of getting a car back on the road.

Tractable already works with

Hartford Financial Services Group Inc.

in the U.S. As well, it said its technology is in use by carriers in Japan, France and Spain.

Industry consultants said the technology works best on photos of vehicles with visible exterior damage, especially fender benders.

Mr. Buffett historically has been slower to embrace new technologies across his operating businesses and investment portfolio. By the time he chooses to invest heavily in new technology, there is usually significant data showing that an innovation has staying power.

One example of this trend: Mr. Buffett for years refused to buy technology stocks outright. He bought his first 10 million shares of

Apple Inc.

in 2016—and owns more than 887 million shares, or over $100 billion worth, as of the end of the first quarter.

Mr. Buffett’s insurer, Geico, was behind rival

Progressive Corp.

in embracing telematics, a way to collect information about mileage and an owner’s driving habits directly from a vehicle. During each of the past two Berkshire annual meetings, Mr. Buffett has addressed Geico’s hesitance and said it was a mistake.

At this year’s meeting, Berkshire Vice Chairman

Ajit Jain

said Geico will make improvements.

“Geico had clearly missed the bus and were late in terms of appreciating the value of telematics,” he said. Geico, Mr. Jain added, has “a number of initiatives, and hopefully they will see the light of day before not too long.”

Geico’s decision is one of many that car insurers are grappling with as more technology becomes available.

The insurers’ move into artificial intelligence is aimed at eliminating common, frustrating experiences for policyholders in filing claims. In the most time-consuming steps under conventional practice, adjusters make appointments to meet policyholders in person to inspect a damaged vehicle, or visit a body shop where the vehicle has been towed. The process can take days to complete.

Write to Leslie Scism at [email protected] and Geoffrey Rogow at [email protected]