Autonomous vehicles are on the road and being tested — and predictions are that we’re moving rapidly towards this reality. We contacted Ron Montoya, senior editor at Edmunds.com and Jeremy Anwyl, an independent auto industry consultant and analyst by email to get their take on the issues.

Motor Matters: Which states will be more inclined to embrace autonomy first?

Montoya: The states that are most inclined to embrace autonomy early are likely to be the ones that have already been the most friendly and accommodating about allowing testing on their roads: California, Nevada, and Michigan.

Anwyl: I would actually see this happening much more quickly in Europe, where governments work much more collaboratively with the European car companies, and the political environment is usually more pragmatic. As autonomous technologies are mainstreamed in Europe, pressure will build for U.S. agencies to get out of the way.

As for the U.S., California probably leads the way. But, for the reason I outlined above, there is a big difference between licensing test vehicles — which still require a human “passenger” — and moving ahead with fully autonomous vehicles.

Q: What are the obstacles in terms of insurance claims?

Montoya: The biggest question about insurance claims is who is at fault if a self-driving autonomous car causes a collision: the person in the driver’s seat or the automaker who built the car? Volvo has said that the automaker should be responsible when the car is in self-driving mode, but not every party involved may take that approach to this complicated issue.

Anwyl: In most states, “no fault” insurance is in place already. Insurance companies, who have very clear thinking when it comes to claims costs and premiums, should be big advocates for autonomous vehicles — at least in the short-term. (As autonomous vehicles reduce the number of vehicles in private ownership, this may also put pressure on insurance companies to reduce costs as their customer base dwindles.)

Q: How long will autonomy take to be say, 20 percent of the U.S. fleet? How far out is the industry looking?

Montoya: It will take decades before 20 percent of the cars on the road are fully autonomous. Automakers may not even offer such a product until around 2030, which is the target date that Kia has specifically named. The first fully autonomous vehicles will be very expensive and likely sold in very low volumes.

We’re still a ways out because, so far, we anticipate that all autonomous cars will be electric vehicles, and their sales make up less than three percent of the U.S. market.

Edmunds.com asked of approximately 1,000 people at a regional auto show: Would you consider buying a self-driving car when one goes on sale? About 50 percent answered, “yes.” The auto industry knows all too well that a lot of people may say yes to an idea, but when you ask them to pay for it, their answers change.

Anwyl: It will happen in phases. Already, we see the second wave of self-driving technologies being introduced. (The first phase were things like ABS and traction control.) Self-parking, lane change warnings, advanced cruise controls; all of these are building blocks towards fully autonomous vehicles. Technically, we could be ready in a few years, but regulators will be among the hurdles.

Initially, customer demand will encourage car companies to develop and introduce self-driving technologies. But a conundrum is developing: compelling alternatives to vehicle ownership. (Autonomous Uber fleets, etc.)

Q: How will legislation roll out?

Anwyl: Big question. In many ways regulation will be the biggest brake on rapid adoption of self-driving technologies. This might seem counter-intuitive, considering human error causes most accidents and the great promise of an autonomous future is a dramatic reduction in traffic-related fatalities and accidents.

So why do I see regulators slowing things down? This gets to the inherent nature of bureaucracies and how they manage political risk. As autonomous vehicles are approved by regulators, the overall fatality rate may go down, but inevitably there will be some fatalities from accidents with autonomous vehicles. (Even if that vehicle was not at fault.) The media will throw a spotlight on these fatalities, and this represents political risk for the regulatory agency. Aware of this risk, the easy way out for a regulatory agency to do as little as possible.

When testing autonomous vehicles today, their accidents have typically been caused by human-driven vehicles crashing into the autonomous vehicle. These accidents weren’t the fault of the autonomous vehicle — or were they? Most of these accidents were a result of the autonomous vehicle obeying the law. We’ve learned that, in the real world, obeying traffic regulations can actually result in a traffic hazard.