While the car was on autopilot when the accident occurred, autopilot is not considered true autonomous technology — driver error on the part of the Tesla driver and/or the driver of the truck it collided with may have played a role in the accident, yet to be determined. While this is not a fully autonomous car fatality, it's close enough that this may be the first event that starts to address shifting liability issues, ultimately impacting how coverage evolves.

One of the first things insurers need to take into account as true, autonomous cars hit the market is that the nature of car ownership may change, affecting the potential client base (number of car owners). Toyota has invested in Uber, General Motors in Lyft, and Volkswagen in Gett, which signals that the availability of alternate transportation options will change people's views toward car ownership. As autonomous vehicles become available, these services, plus others such as ZipCar's car-sharing programs, will be able to deploy resources to underserved markets.

Impacts will be significant on costlier insurance markets, such as urban areas, as there will be noticeable decreases in full-time car ownership. This will also have similar, but likely smaller, impacts to existing markets in less densely populated suburban areas. With the availability of reliable and affordable alternate autonomous private car services, a two-car household may reduce to a single-car household augmented with ride-sharing or car-sharing services, for example.

As autonomous cars become reality, liability may shift from the driver, which will result in multiple changes to insurance products. Multiple automakers, with Volvo leading the charge, have stated they will accept the liability for their cars while in autonomous mode. Volvo's statement was the most sweeping, accepting full liability for autonomous mode driving, while Mercedes and Google will accept liability when their technology is at fault; these two liability assumptions may sound similar, but the Mercedes/Google assumption of liability is narrower and open to interpretation, unlike the Volvo assumption which calls for assuming "all liability." This indicates that, to a certain extent, the liability portion of the auto policy will become more of a warrantee issue.

If liability is not assumed under warrantee, loss may become a product liability issue. Injured parties, or insurers through subrogation, may seek damages directly from automakers/technology provider, which would most likely trigger a product liability claim. The impact on insurance products (and associated legislation) is far from decided. Insurers can expect to see some major changes in the following product areas:

• Liability issues will dramatically change the exposure, which will lead to changes in underlying auto liability coverage. How much is shifted to warrantee/product liability remains to be seen, and the role of legislation will be significant. In the beginning, expect the status quo, but insurers should be prepared to change on a dime. Personal lines insurers may decide to venture into quasi-commercial coverages to support the emerging market, and commercial lines insurers will at a minimum need to identify new exposures under traditional products (product liability) or create new products (product liability and warrantee).

• Comprehensive/collision coverage impacts would be minimal, with these products staying relatively similar. Frequency and severity of losses should go down, along with premiums, as cars themselves become safer and losses will be due more to environmental damages such as severe weather (hail) or falling objects (tree limbs).

• Subrogation will be an interesting topic to watch as liability issues work themselves out in the market. Eventually, standard subrogation norms will be established, and insurers should be prepared to optimize their subrogation processes throughout the maturation of the new auto market. This will include taking advantage of market confusion as the autonomous car emerges.

• Micro insurance products may see a surge, covering aspects of auto usage with regard to usage of car share and autonomous car programs. These products would be point of sale transactions, tied to a rental or account usage.

• Pricing will significantly drop for auto insurance, insurers must be prepared to respond to this change from both a product and an operations perspective. Elements used for pricing products today (such as driver criteria) will change tremendously. This sounds like a small issue, but most insurers would have a hard time changing pricing models today.

With this in mind and autonomous cars just over the horizon, what do insurers need to do to get ready? The answer lies in five key areas:

Related: BMW to deploy fleet of self-driving 7-Series on roads this year

 The self-driving Tesla Model S sedan. (Photo: P.Harman/Claims magazine)

Model trending on current safety improvements

Every model year release includes new safety features that make cars safer. These innovations aren't a progression toward the emergence of the autonomous car, but they do provide data points and trends that insurers should be monitoring and leveraging in pricing models for future autonomous car insurance pricing. Keeping in mind that this pricing evolution will be a step-by-step process, but by monitoring the emerging frequency and severity trends associated with emerging safety technology, insurers should be able to model future pricing much better than starting from scratch when the new market emerges.

Strategic scenario planning

How the insurance market, product, pricing and regulations for autonomous auto insurance will change and develop is up in the air right now. Additionally, autonomous auto efficacy and adoption is still unknown. With all this uncertainty, insurers should adopt a scenario-based strategy process for responding to the market. Leveraging scenarios based on different market outcomes, insurers can have operational, product and IT strategies in place, ready to execute based upon market characteristics as they develop.

Flexible product environment

In standard markets that are not prone to huge shifts, insurers find getting new products to market quickly and effectively a major hurdle to their business success. This hurdle is magnified in a redefined and shifting market. Insurers need to invest in systems that will enable them to get new products institutionalized quickly. This doesn't mean just filing with insurance regulators, but being able to update underwriting, rating, quoting, front-office systems, and policy administration systems quickly and easily.

Most insurers don't have the time or the investment appetite to do wholescale replacement of systems. Instead, a wrap and renew process, where a service layer that supports users but separates the underlying legacy systems, will work in creating a responsive environment, in essence a bi-modal IT environment.

Flexible operational environment

Getting a new product filed is one thing. Selling and servicing it is a completely different issue. Insurers need to be able to provide user platforms (direct sales, captive and independent agency sales, as well as service) that will enable them to sell and service the client without disruption as products, underwriting, and processing changes. This includes pushing information to the agents and customer service representatives on how to position new products effectively without taking them away from their work to train them every time a change comes about.

An insurer should have two goals when preparing for the new auto market: minimize the disruption for the user on the sales and service side, and automate processes as much as possible. With the expected lower premium base and uncertain results, processing expenses need to be kept to a minimum with straight through processing as the norm wherever possible.

Actionable analytics

This may be the most important point. Insurers will need analytics to help them drive their strategy and reaction to the market as it develops. Winners in the upheaval produced by autonomous cars will not be the first insurers with a tailored product to market, but the insurers able to monitor the development of this reformed market and respond quickly and effectively to how it develops. This means tying analytics directly to underwriting and claims operations, watching how loss and expense ratios develop, and being able to change the go to market strategy immediately.

This could entail changing pricing, underwriting guidelines, claims processes, market segmentation, and more as results develop. Implementing actionable analytics will provide insurers with insight into how the business develops and the ability to pinpoint where results are not meeting expectations, highlighting where changes needs to be made. Analytics should be tied very tightly to scenario-based planning, leveraging established plans prepared before the market shifts, instead of having to respond on the fly.

To paraphrase Louis Pasteur, fortune favors the prepared. There is no doubt there will be upheaval in the auto insurance market. How much change and how intense it will be remains to be seen. Being early to the market with prepared products will always give an insurer first mover advantage. However, the ultimate winners in the marketplace will be the insurers that plan strategically and have nimble, efficient operations that can respond quickly and effectively to market changes.

Tom King is the senior director and industry principal of insurance at Pegasystems (www.pega.com). King has more than 25 years of experience in the insurance industry, and much of his career has been dedicated to helping insurers leverage technology in conjunction with defining business strategy.