Autonomous Cars

Self-driving cars present some intriguing possibilities to me, as a piece of technology.  I personally like the idea of reading a book while my car drives me to work in the morning, and some of the ownership possibilities (such as owning a self-driving car with a small group of people) are especially tempting.  Google famously has a fully autonomous car in the works, but other companies also have their own self-driving cars in the pipeline.  These cars run from being fully autonomous to having certain features in the car automated.  For example, Volvo has a braking system that automatically detects obstacles on the road and adjusts accordingly. Similarly, Mercedes invested tons of money into a system that can stop the car or keep it in the lane in an emergency as an additional feature for their S-Class sedan.

As a geek, I find this technology fascinating.  As an attorney, I wonder about how the legal system will treat these vehicles.  Currently, establishing liability for speeding or accidents is pretty simple.  Cars have an owner and operator (though they aren’t always the same person) who would possess some liability should the car be involved in some violation of the law or cause injury.

First, certain common driving violations might not even apply to self-driving cars.  There are a number of driving laws designed to combat certain human limitations. Speeding and DUIs represent the most obvious examples of this category of law.  A computer operated vehicle does not get drunk (though if it could, that would represent a rather strange breakthrough), rendering the DUI inapplicable.  Speeding is built around limitations in human reflexes leading to state-issued determinations on what speeds in which it is safe to operate a car.  Computer technology not only gets around some of those limitations, it can likely be set to adhere to speed limits if need be.  There are even other laws that might go by the wayside.  Running a red light, for instance, might not be necessary since intersections would operate differently.  As this article from The Atlantic demonstrates, a world of self-driving cars could handle infinitely more complicated intersections since this system ensures a constant flow of traffic based on what everyone else is doing (no lights necessary).

That leaves the issue of tort liability, which will likely still exist.  It is not clear how much legal responsibility the human driver bears for the actions of the vehicle, since one of the assumption with self-driving cars is that the human operator can assume manual control of the car in situations the computer cannot handle.  As this Stanford Center for Internet and Society white paper points out, owners often retain some liability for the actions of a person whom they let drive.  There would still be a need to operate the vehicle prudently (as currently required by law), and a driver may still be liable for negligent actions committed during operation of a self-driving car.  The issue would be when that liability begins: would it begin when an individual directly operates the car, would they need to intervene to stop an accident during certain circumstances, or would the owner of a car always remain liable for any accidents resulting from operation of the self-driving car?

A significant part of determining liability also hinges on how states define the prudent operation of a motor vehicle in this circumstances.  The Stanford article claims that this arises in three ways: inputs (the instructions the human operator provides to the car), outputs (how the vehicle performs with other cars), and vigilance (how much attention the driver must pay to the road).  All three provide some duty on the part of the driver and thus creates the possibility of negligence.  However, state determinations will determine the precise nature of these duties and how they impact drivers.  Requiring a human driver to pay as much attention to the road in a self-driving car as they would when driving manually would impose a potentially high duty to pay attention even when one does not actually drive the car.  Self-driving cars will likely get into accidents less frequently than people (given how big of a factor human error represents in-car accidents), but an owner may retain some responsibility for the accidents that do occur due to this prudence requirement.

Finally, Bryant Walker Smith points out another, more subtle issue in this article from Slate.  Self-driving cars have the potential to collect a great deal of data regarding the comings and goings of their owners.  How companies use and secure that data (and when they’re forced to disclose it) represents a major question.  In order to operate, these cars use a variety of positional sensors, GPS, and driving data that would tell those reading the data an extraordinary amount about the person who owns the car.  Think about how much someone could tell about you if they could see exactly where your car went over the course of a few weeks or months.  I would go further than Smith, and wonder about the Fourth Amendment implications of such a technology.  Would courts consider an owner to have a reasonable expectation of privacy in the data held by their car?  Would police be able to conduct a search of such data without a warrant?  Courts would likely have to spend years considering the expectation of privacy people have in the data stored in a car’s onboard computer?  The closest case I can think of is U.S. v. Jones, which dealt with the police placing a GPS tracking device on a car without a warrant.  The court found that use of such a device to track a vehicle’s movements represented a search under the Fourth Amendment (and thus needed a warrant).  However, the majority of the Court argued that placing the GPS tracker on the car represented a trespass (which would not apply to a computer in the car itself).  Two other justices wrote opinions articulating different theories (one saying that the GPS tracker violated the defendant’s privacy and the other questioned the general constitutionality of GPS surveillance).  Until the Supreme Court hears a case dealing with a police search of the information stored on a car’s onboard computer, there likely won’t be a definitive answer to this Fourth Amendment question.

How states and courts resolve these issues will take years of policy and legal debates to sort out.  While the technology may be coming quickly, the changes a self-driving car represents for transportation regulation are enormous.  I still look forward to seeing how these issues get resolved, as well as seeing one of these cars in action for myself.

I also have a small update.  In observance of Rosh Hashanah (the Jewish New Year), I will not have a post for Friday.  Enjoy the rest of the week everyone.


2 thoughts on “Autonomous Cars

  1. I’m just as interested in how the legal system will deal with self-driving cars, since it is a completely new field. My own bet is that we’ll eventually see autonomous taxis roaming the streets, picking up and dropping off passengers. Such fleets of “robo-taxis” would likely be operated by companies, which brings business law into the picture as well. I wrote about them on my own blog a couple weeks ago (, but I didn’t touch on the legal part of the subject, as I don’t know much about it.

    I imagine the software in charge of the car could be another major liability question, too. How much liability would developers have for the cars their software is operating? How about security to protect against hackers? And what about third-party modifications and plugins? (Because someone will come out with an open source version eventually…) This is weird, weird stuff… now you have me thinking!

    Out of curiosity, do you believe changes in the law will keep pace more or less with the development of technologies like this, or will it end up retarding technological progress? It’s something I’ve wondered about.

    • Generally, it depends on the technology. For faster moving segments of the tech world (such as internet based technologies), I’m generally of the opinion that the law struggles to keep pace with new technology. Passing new bills just takes too much time, especially compared to the development cycles in certain industries (which leaves whatever lawmakers passed a generation behind). Judges, who are better equipped through the power of precedent to respond more quickly to technological changes, often have a murky understanding of science and technology. Occasionally a court does a good job informing themselves about how a particular technology works when coming to a ruling, but that is definitely the exception. These issues seem to subside if previous law and precedent covers the new technology well.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s