How would you feel if 40% of your business would be at risk of slowly melting away? That’s exactly the situation most property and casualty (P&C) insurance companies are facing. Autonomous driving is despite all the social benefits an obvious threat to the bread and butter business of insurers.

I take it for granted that fifty years from now we’ll look back and wonder How on Earth did we allow humans to drive cars for so long? It's reasonable to assume that in the future most vehicles on land and sea, probably also in the air, will be controlled by robots and artificial intelligence.

I’ve written several blog posts about the advent of robots and artificial intelligence (see for example “Artificial Intelligence – our final invention"or “When robots do all the work, what should we do?"), so in this blog post I’ll only discuss autonomous driving.

Recently Richard Holman, a 30-year automotive veteran running GM’s foresight and trends unit, said that three years ago most industry participants would have estimated 2035 as a reasonable timetable for self-driving cars. Thanks to Silicon Valley companies like Alphabet's Google Car Division and Tesla Motors, he now estimates that most people see self-driving technology being deployed by 2020, if not sooner.

It’s not just Google and Tesla that are increasing their efforts on self-driving cars. Every major car manufacturer has woken up and and R&D spending on autonomous driving is exploding. BMW is going to ramp up hiring so that half of its R&D team is made up of software engineers. They will be tasked with building artificial intelligence, which will power self-driving BMW vehicles of the future.

Volvo points out that already now their latest models use semi-autonomous technology to make life easier. They monitor the road ahead, alert you to danger and, when necessary, help out with steering and braking.

Volvo is so confident in its efforts that it claims it will take the blame if one of its self-driving cars crashes. In essence, this is Volvo’s response to the slow-moving regulatory framework, which needs to capture how liability works in an self-driving world. The key question has been who takes the blame when a car controlled by a computer gets into a crash – and Volvo has just given us a bold answer.

So how is improved safety bad for insurance companies? Aren’t the technological advances beneficial for just about everybody from a social perspective? Yes, and in fact insurance companies have always been proponents of safety. But as the risk of accidents diminishes, so does the insurance premiums because the business of taking on risk is what insuring is all about.

There will be many reasons for the decrease in motor insurance premiums. First and foremost there will be a lot less traffic accidents, which means a lot less people injured and cars wrecked. Diminished risk leads to lower insurance premiums. Second, auto thefts will probably also decrease substantially due to improved technology and tracking. In the future you need to be a super-hacker if you want to steal a car. 

It may also be that the regulatory framework changes completely on the liability part of motor insurance. What if it becomes the auto manufacturers' responsibility to take on the liability risk, just as Volvo is anticipating? Will the manufacturers carry that risk completely or reinsure parts of it? Whatever the regulatory changes may be, it seems clear that there will be changes in required levels and types of motor insurance coverage.

Like with all major game-changing trends the drop in premiums will not occur over night but gradually. Nevertheless it will be a dramatic change. Looking at the largest insurer in the Nordic region of Europe, If P&C Insurance Company, Motor and Motor Third Party Liability (MTPL) accounted for 42% of the Gross Written Premiums (GWP in 2015 was 4,6 billion euros). Motor insurances are also very profitable for insurance companies in general. For If P&C Insurance Company Motor and MTPL insurances combined generated a technical result of 212 million euros in 2014 and 143 million euros in 2015.

I assume that the distribution of premiums and profits is similar for most European insurance companies. The situation is probably similar in the U.S. as well. Recently Warren Buffett, the Chairman and CEO of Berkshire Hathaway, said that "in the long term insurers are going to be bringing in less money as self-driving cars start to take over”. He also  concluded that “anything that makes cars safer is very pro-social, and it's bad for the auto insurance industry”. One of the cash cows of Berkshire Hathaway has been the wholly owned auto insurer GEICO.

Fortunately for insurance companies, the motor insurance business will probably not melt away completely. For example, autonomous driving won’t remove the risk of natural disasters, like floods and storms. And who knows what new risks will emerge in the future as we’re entering a robotized society. Where there are new risks, there will be new insurance products.

The insurance companies are probably well aware – and worried – of the premium risks involved in self-driving cars. Torbjörn Magnusson, CEO of If P&C Insurance Company, wrote in the 2015 annual report that “smarter technology in cars and other things surrounding us will affect the way we design our insurance products”. I'd call that a classical example of converting a challenge into an opportunity.

However, I believe the challenge is much more grave to the insurance companies than what they publicly want to admit to owners and stakeholders.