A recent Wall Street Journal article1 opines that self-driving cars have “crossed an important milestone,” because three automobile insurers recently stated, in annual reports and related communications, that the technology might “materially alter the way that automobile insurance is marketed, priced, and underwritten.”
The questions are real. When a self-driving car (either “driver” occupied or otherwise) is involved in a collision, and injures either a passenger, an occupant of another vehicle, or a pedestrian, who is (or should be) legally responsible for damages in California?
Should the answer differ if the vehicle’s “driver” maintained some mechanical, discretionary control over the car’s operation? Should all liability shift from the vehicle’s “driver” or owner to the manufacturer instead; or should that occur only if there are proven “defects” in the car’s self-operation apparatus, rather than just a single incident of negligent operation?
Thoughtful law review articles have pondered the problem. One finds a potential solution in application of the tort doctrine of strict liability to the vehicle owner, making legal responsibility akin to that one bears, as owner, for the conduct of his or her dog, or for the possession of hazardous or explosive materials.2 Another suggests the possibility of legislative limitation of liability to support this socially beneficial technology (apparently leaving plaintiffs uncompensated for their injuries); or, alternatively, federal preemption of state tort actions otherwise available to an injured plaintiff, so as to protect the fledgling driverless car industry.3 Still another suggests differentiating the respective legal responsibility, and resulting liability for injuries to plaintiffs, according to categories of vehicle users: attentive drivers, distracted drivers, diminished capabilities drivers, and disabled drivers.4
Another possibility would be to assume the adequacy of current law: that is, that injured victims harmed by driverless cars will be compensated by product liability claims against the vehicle manufacturer. But this assumption has pitfalls as well.
“In a products liability case involving a motor vehicle, as in other products liability cases, the protective ends of the law are ‘attained by the necessity of proving that there was a defect in the manufacture or design of the product and that such defect was a proximate [or legal] cause of the injuries.’” Doupnik v. General Motors Corp. (1990) 225 Cal.App.3d 849, 859 (internal citation omitted). Many collisions involving driverless vehicles may occur absent any provable defect in the vehicle’s operative system. A court might find that just because an operating system of a driverless vehicle is not perfect, and capable of accident-free operation on a daily basis, does not mean it is provably “defective.” Cf., Dabush v. Mercedes-Benz USA, LLC (N.J. Super. Ct. App. Div. 2005) 378 N.J.Super. 105, 122, rejecting a user’s claim of alleged deficiencies of an automobile navigation system because: “The navigation system is exactly what it was designed and intended to be; an aid to navigation, not a perfect instrumentality of navigation.” So relying on product liability law as the sole potential source of an injured plaintiff’s recovery from an accident involving a driverless car is not realistic either.
Finding a legislative solution is not impossible. State legislatures do adapt to changing times, customs, and technologies. For example, effective July 1, 2015, AB 2993 goes into effect in California, adding Sections 5430 et seq. to the Public Utilities Code. These statutes, which ultimately obtained the blessing of most industry “players” prior to passage, establish tiered limits of liability for drivers while they are operating their personal vehicles for so-called “transportation network companies” (like Uber and Lyft).
The liability limit in force for the network’s driver at a given point in time will depend on whether she is awaiting a dispatch to pick up a potential passenger; or whether she is actually on the way to pick up the passenger; or whether the passenger is already onboard. While legal and legislative adaptation to this new industry model of “carriage for hire” is far from over (e.g., the status of drivers as independent contractors or employees remains the subject of extensive litigation), a dilemma that seemed unsolvable just months ago has now been dealt with legislatively in a responsible manner. Can that be accomplished for driverless cars as well?
In fact, without much fanfare, the legislature started to deal with driverless cars over two years ago. Vehicle Code § 38750, passed in 2012, addresses insurance, monitoring, reporting, public hearing, potential regulations, and other requirements for testing autonomous vehicles; and it provides that they may be operated on public roads so long as, per subdivision (b)(2): “The driver shall be seated in the driver’s seat, monitoring the safe operation of the autonomous vehicle, and capable of taking over immediate manual control of the autonomous vehicle in the event of an autonomous technology failure or other emergency.”
That statute presents a starting point for regulatory control of self-driving vehicles that might also encompass an expansive resolution of liability and insurance issues. Hopefully this can be accomplished by the time the safety and practicality for daily use of these vehicles has been demonstrated. But the statute, as written now to provide for vehicle testing, gives little guidance as to what might constitute an overall scheme setting out who will be liable in tort to an injured plaintiff.
What might that scheme be? While far from perfect, California’s tort law is a functioning judicial system; and it is one unlikely to change much, legislatively, over the next few years. It is our hypothesis that, in the short term, the preferential tool for dealing with liability and compensation issues arising from the use of driverless cars would be one most in tune with, and least disruptive of, the civil justice system’s current operation.
One way to implement that goal would be an amendment like this one to Vehicle Code, § 38750: “The operation of a vehicle, in an autonomous or semi-autonomous mode, shall be imputed to (and be deemed the conduct of) its primary operator.” A statutory definition of “primary operator” would be added to assure that it encompassed, where applicable, a non-driver occupant (or a non-occupant) who, while owning, controlling, or permissively using the vehicle, established and entered the trip’s destination into the vehicle’s operating system. If such a statute were enacted, the statutory limited liability (set forth in Vehicle Code Section 17150) of a vehicle owner for negligence while her vehicle is being permissively used by another could remain unaltered.
Nor would other existing law need to be altered: the rights, obligations, and remedies of litigants could remain in force. When counsel sees fit, an injured plaintiff would be free to set out an additional cause of action against the vehicle’s manufacturer for product liability; or a plaintiff could choose to sue just the manufacturer (not the “driver”). Moreover, an innocent vehicle “operator” or owner sued by plaintiff could file a cross-complaint against the manufacturer, seeking indemnity for any losses he might incur as a result of plaintiff’s claims.
If the self-driving car is determined to have been proceeding with due care (non-negligently), there would be (as there is now) no legal liability on the part of its owner or operator for the plaintiff’s claims; otherwise, the owner and primary operator of the driverless vehicle (usually, through her insurance carrier) would compensate an injured plaintiff for his damages, just as a negligent driver, and his insurer, would have to do now.5 Fault would be litigated just as it is now.
With the passage of time and the accumulation of claims experience involving driverless vehicles, perhaps greater alterations of the insurance and tort liability laws will be deemed appropriate. But the minimal changes suggested here should present an interim solution that is consistent with both fairness and the existing legal responsibility a driver has (pursuant to Civil Code Section 1714) to exercise ordinary care in the management of his or her property. If one chooses to use driverless technology, it is not unreasonable to bear legal responsibility for negligence that occurs during its operation.
Automobile insurance rates for the owner of a driverless car will reflect both those risks and the technology’s potentially good safety experience; and issued policies would (as they do now) protect policyholders at the purchased level of applicable policy limits. One would expect liability insurance rates to go down, as the technology of driverless cars improves and as they became more prevalent, since driver error accounts for a large percentage of current vehicle accidents.
The legislature should start to deal with these issues now. Preemptively passing legislation that anticipates widespread use of driverless cars in the future would permit insurance carriers to begin to evaluate risks, and start the process of underwriting insurance policies that anticipate such usage.
Further thoughts on liability and compensation issues for ‘self-driving’ cars
In a recent blog article, we suggested that the preferential tool for dealing with insurance, liability and compensation issues arising from the use of driverless cars would be one most in tune with, and least disruptive of, the civil justice system’s current operation.
We noted that one way to implement that goal would be an amendment like this one to Vehicle Code, § 38750: “The operation of a vehicle, in an autonomous or semi-autonomous mode, shall be imputed to (and be deemed the conduct of) its primary operator.” A statutory definition of “primary operator” would be added to assure that it encompassed, where applicable, a non-driver occupant (or a non-occupant) who, while owning, controlling, or permissively using the vehicle, established and entered the trip’s destination into the vehicle’s operating system. If such a statute were enacted, the statutory limited liability (set forth in Vehicle Code Section 17150) of a vehicle owner for negligence while her vehicle is being permissively used by another could remain unaltered.
Dealing with these issues now seems especially timely, given their recent prominence in the news. The latest headline is in the San Francisco Chronicle dated May 12, 2015. It states: “Denting Their Credibility: Google Self-Driving cars Have 11 Minor Crashes in 6 Years.” In the true spirit of California’s adversary civil justice system, the Google representative’s quoted comment denies fault for any of these incidents: “Not once was the self-driving car the cause of the accident.”
Since Google and other experimenters are the owners of the current 48 driverless cars on California highways, issues of liability and compensation in these incidents can be investigated and adjudicated, as needed, under existing law.
But before these vehicles do get more universally into use by the driving public, the insurance, liability, and compensation issues mentioned above (and explored at length in my prior article) must be addressed. The primary legislative and judicial concern will be whether a vehicle owner should bear liability for the negligent operation, if any, of a driverless car. Since there is ample legal precedent cited in my earlier article for the proposition that an owner bears legal liability for malfunctions of her property, this is the legislative course which we suggest.
2. Sophia Duffy et al., “Sit, Stay, Drive: The Future Of Autonomous Car Liability,” 16 SMU Science And Technology Law Review 453 (2013).
3. Gary E. Marchant et al., “The Coming Collision Between Autonomous Vehicles And The Liability System,” 52 Santa Clara L. Rev. 1321 (2012).
4. Jeffrey K. Gurney, “Sue My Car Not Me: Products Liability And Accidents Involving Autonomous Vehicles,” 2013 U.Ill.J.L.Tech. & Pol’y 247 (2013).
5. One objection to such a rule might be that, where the “driver” or owner of the driverless car did nothing negligent (just the driverless car did), the law should not find the “driver” legally responsible. But see Restatement (Third) of Agency §1.04, cmt. e (2006), noting that: “…an inanimate object may serve as a person’s instrumentality in a manner that may be legally consequential for the person…. At present, computer programs are instrumentalities of the persons who use them.” Imposing liability on the “driver” here would be consistent with that ongoing implementation of technological changes into the legal system. See also, discussion on this topic in Duffy, 16 SMU Science And Technology Law Review at pp. 462-467.
The author thanks attorneys Malcolm Wheeler, David J. Samuelsen, and Kelly Balamuth for reading, and offering comments upon, drafts of this article.