Robocar accidents (and AI and robotics in general) bring a whole new way of looking at the law. Generally, the law exists to deter and punish bad activity done by humans who typically knew the law, and knew they were doing something unsafe or nefarious. It is meaningless to punish robots, but in punishing the people and companies who make them, it will likely be the case that they did everything they could to stay within the law and held no ill will.

If a robocar (or its occupant or maker) ever gets a correct ticket for violating the vehicle code or other law, this will be a huge event for the team who made it. They'll be surprised, and they'll immediately work to fix whatever flaw caused that to happen. While software updates will not be instantaneous, soon that fix will be downloaded to all vehicles. All competitors will check their own systems to make sure they haven't made the same mistake, and they will also fix things if they need to.

As such, all robocars, as a group, will never get the same ticket again.

This is very much unlike the way humans work. When the first human got a ticket for an unsafe lane change, this didn't stop all the other people from making unsafe lane changes. At best, hearing about how expensive the ticket was will put a slight damper on things. Laws exist because humans can't be trusted not to break them, and there must be a means to stop them.

This suggests an entirely different way of looking at the law. Most of the vehicle code is there because humans can't be trusted to follow the general principles behind the code -- to stay safe, and to not unfairly impede the path of others and keep traffic flowing. There are hundreds of millions of drivers, each with their own personalities, motives and regard or disregard for those principles and the law.

In the robocar world, there will probably not be more than a couple of dozen distinct "drivers" in a whole nation. You could literally get the designers of all these systems together in a room, and work out any issues of how to uphold the principles of the road.

Much more here on the legal complexities surrounding self-driving cars. I find the topic more interesting than, say, the social implications of Google Glasses. More important, too.