Self-driving cars are becoming the wave of the future. The most advanced models may be ready for launch within two years, despite the fact that they still have the crash record higher than that of a novice driver. The question is thus increasingly posed of who is liable for car accidents involving self driving vehicles. The most likely projection is that cases of product liability and commercial liability will increase as personal liability decreases, but the question of self driving car accident liability could prove to be legally quite complex.
Under the current legal framework, liability is covered under tort law and product liability law. Which type of law applies depends on whether the driver or the car manufacturer is at fault.
Torts are used in cases of personal liability for an accident that causes significant injury, distress or property damage. There are several theories of tort law which may affect the interpretation of fault in car crashes in a variety of ways. Under traditional negligence, no-fault liability and strict liability, drivers in serious accidents would be held accountable in various ways and pay in various different ways.
The upshot, though, is that with the existing technology, over 90% of car crashes are the result of human error for which drivers may be held accountable through tort law. Typically, verdicts and settlements involve payment through car insurance.
Crashes that result from some fault on the part of the manufacturer, on the other hand, are governed through product liability law in terms of negligence and strict liability. Product liability may be found in cases of manufacturing defects, design defects and failure to warn.
Manufacturing defects occur when a product fails to meet the standards and specifications laid out by a manufacturer. Design defects are cases where an accident could have been avoided through an alternative design, and failure to warn cases arise when the maker doesn’t properly advise consumers about how to use a product safely, or about the risks of use.
Autonomous Vehicles and the Law
Recent research by the Insurance Institute for Highway Safety shows that automated features such as forward collision warnings and automatic braking, which are part of currently available Advanced Driver-Assistance Systems, have substantially reduced car accidents. This suggests that, at some point in the development of the technology, autonomous vehicles could substantially reduce car accidents and consequent injuries and fatalities.
This provocatively points to a possible near future in which car insurance costs fall to the point that traditional vehicle insurance itself ceases to be necessary. It is conceivable that car accidents might one day be covered under health and homeowner’s insurance, just as bicycle accidents are. Under such a system, accident victims would be compensated relatively swiftly without the need to identify an at-fault party.
Car companies will hopefully to some extent have an incentive to reduce the defects in their products out of fear of liability. At any rate, under product liability law, a plaintiff will have to demonstrate a defect in design or manufacturing or a failure to warn.
This raises the serious question of how to determine responsibility for software bugs. No court has yet settled any case of a vehicle accident in which software was at fault, and both the outcome of such a case and the precedents it will set for future cases are unclear. There will also likely be a problem finding qualified experts to give testimony in cases like this.
There are four potential defendants in an autonomous vehicle case.
- Vehicle Operator– It’s not entirely clear on the face of it who this will be. Is it the person who controls the vehicle, even if remotely? Or the person sitting in the driver seat? Different states are drawing up different laws on the subject.
- Manufacturer– If the manufacturer played a role in installing the autonomous technology, they could conceivably be held liable, though states like Florida are drawing up laws to limit manufacturer liability.
- Tech Company– The company that produced the autonomous vehicle software could be named, as could the vendor that manufactured the sensors. This is the most likely target of an autonomous vehicle lawsuit.
The technology company will have a range of defenses against a lawsuit of this sort. In comparative negligence, the company could argue that the vehicle operator shared in the fault for the accident. For product misuse, the company could argue that the operator may bear responsibility for ignoring instructions regarding the vehicle’s proper use, or may have altered it in some way that rendered it hazardous. Finally, a state of the art defense could argue that safer design alternatives did not exist when the vehicle was made.
Hacking and data breaches open up another legal arena. Whether personal data is stored in a car or in the cloud, manufacturers and software vendors could be held liable for information leaks as well as for any accidents that result. The same holds true for the transportation grid as a whole.
Artificial Intelligence and the Law
Software that can manipulate physical things such as vehicles and robots in the manner of an advanced artificial intelligence program has the potential to cause property damage, injury and death, and as such is subject to criminal and civil liability.
Experts have identified three ways that responsible parties may be identified and held liable for damages arising from faulty software.
- Perpetrator via Another– If an individual, probably either the programmer or operator, directly tells an AI to commit a crime, they can be held accountable under this principle. In current law, it is used when someone instructs a legally incompetent person, such as a child or severely mentally disabled person, to commit a crime.
- Natural and Probable Consequence– If the likely outcome of one of the AI’s protocols is to inflict damage against person or property, the programmer can be held liable. If an AI were programmed to run over a pedestrian who crossed against the light, for instance, this principle would apply.
- Direct Liability– Strict liability in certain spheres of law, including traffic law, requires action, unlike the intention required in, for instance, felony criminal law cases. The AI system itself might be assigned criminal responsibility under the direct liability principle, but it’s not clear what purpose this would serve.
Yet another question that bears on self driving car accident liability is the likelihood of major shifts in policy and legislation on the topic in the near future. This has substantially already begun, with public polity debates on the topic and new laws being drafted at the federal and state levels.
The National Highway Traffic Safety Administration released a report in September of 2016 on the progress of the self-driving vehicle and the regulatory changes needed to accommodate it. They determined that the states would bear primary responsibility for legal regulation of autonomous vehicles and that, in turn, decisions made at the state level would have a substantial impact both on how quickly autonomous vehicles get rolled out and on insurance costs in the states.
The US Congress addressed the issue in the SELF DRIVE Act of 2017, which reaffirmed the federal government’s role in regulating product safety and formulating liability laws. This piece of legislation signaled that at least some of the states’ current responsibility for regulating civil liability would be shifted to the federal level, though it left aside the question of civil and criminal liability for manslaughter.
Many American states have already begun drafting legislation adjusting civil liability in the case of autonomous vehicle accidents. For example, Nevada passed a bill shielding from liability the manufacturer of a vehicle converted into an autonomous vehicle by another company. Michigan and Washington, D.C. have passed similar statutes.
There are mounting concerns that the shift in legal and insurance liability from the tortious to the commercial will result in car makers and software companies overbearing the costs. That could lead to higher costs and needless delays in the rollout of the new technology. Numerous proposals have been floated to mitigate this problem.
Conclusion: Who is liable in a Self Driving Car Accident?
The question of who is liable in an accident involving a self-driving car is currently in flux.
Although it is clear that, under existing law, a fully automated vehicle will shift virtually all of the legal burden for crashes from the driver to the manufacturer and from torts to product liability, it is not clear whether the laws will stay the same or if self-driving cars will have any future if the laws don’t change.
Manufacturers may be so nervous about assuming sole liability that they delay the commercialization of the technology until it is literally perfect, and the day on which that happens may never dawn. More likely, federal and state law will shift to accommodate the new tech, but what the new legislation will look like remains unclear. We can do little but hope that the new world of transportation that appears to be waiting off the next exit ramp will be safer and more convenient.
If you find yourself in a legal battle, it’s always best to consult a car accident lawyer.