Top>Opinion>Robot-Car Liability
Susumu Hirano [profile]
Susumu Hirano
Professor of Law, Faculty of Policy Studies, Chuo University; Dean, Graduate School of Policy Studies, Chuo University
Areas of Specialization: American civil law (torts/contracts), cyber law (Internet law)
Autonomous vehicles (robot cars) have suddenly started to attract worldwide attention. Those designed by the search-engine giant Google have equipment that measures the environment around the entire periphery of the vehicle using sensors mounted on the roof. Reportedly, the acquired information is then compared with map information previously stored in data centers, and the vehicle’s accident-free record for driverless, autonomous travel is updated. Three states—Nevada, Florida, and California—have already enacted statutes permitting the operation of [quasi-] autonomous vehicles on public roads and are encouraging the practical application of this technology.
Autonomous cars offer significant benefits. For example, (1) since most car accidents are caused by human error on the part of drivers, autonomous cars’ elimination of the need for a driver can greatly reduce accidents. The cars can also (2) save people time that they have traditionally spent on driving. Furthermore, (3) by enabling extremely efficient driving and avoiding wastage, they can curb the loss of fuel resources and, by extension, reduce environmental degradation. In addition, they would (4) allow the old, the young, and the disabled, who have been barred from the benefits of mobility due to their inability to drive, to enjoy these benefits.
While autonomous cars offer many benefits, they also raise product liability concerns for manufacturers and other parties due to their autonomous nature. Even autonomous cars cannot prevent all car accidents. Once an autonomous car causes an accident, however, 1) the responsibility cannot be attributed to driver error, so there is a strong likelihood that the accident will be viewed as the result of a malfunction or defect. In other words, the responsibility/liability is shifted from the driver to the manufacturer and other parties. In addition, 2) if vehicles that are supposed to autonomously (and, of course, safely) arrive at their destination cause accidents due to malfunction, the entire product category could be labeled as dangerous and defective under “product-category liability.” This liability could then 3) lead to widespread criticism and condemnation of the manufacturers and other parties that produce autonomous vehicles. There is an unshakable concern that the resulting negative reputation could, for example, cause boycotts of other key products offered by the same companies, putting their very survival at stake.
The three concerns described above are not wholly unfounded. 1) Product liability precedents in both the U.S. and Japan show a trend—called the “Malfunction Doctrine” or “factual presumption (defect ipsa loquitur)” —to readily hold manufacturers and other parties responsible in the event of an abnormal accident due to malfunction.
2) Based on product liability precedents in the two countries, however, we can also make a rough prediction that this would not easily extend to product-category liability, which categorically treats all autonomous vehicles as defective. Especially in the U.S., the courts (the judicial branch) have been analyzed as being conservative in their use of authority to treat an entire product category as defective and bar it from the market, with a tendency to leave such highly influential policy decisions to the executive and legislative branches of government that have jurisdiction over product safety. There is no guarantee, however, that the executive and legislative branches of government will not make a move to put restrictions on the autonomous vehicle product category. This can be easily surmised from the U.S. government's response to a former alleged defect in Toyota Motor's hybrid cars.
Finally, 3) the Snow Brand food poisoning incident elegantly demonstrates that the fear a company’s “negative reputation” could lead to boycotts and drive it into bankruptcy is not unfounded. Since there are fewer lawsuits in Japan than in the U.S., however, some argue that the potential losses are negligible. Even so, it is not the financial expenses of lawsuits that companies are truly worried about, but the intangible effects of a negative reputation. It has been pointed out that new products are more likely to be viewed as risky to an excessive, irrational degree, due to their enigmatic novelty, than products that have been in widespread use for a long time. New products attract people’s attention more readily than conventional products. And since the autonomous car is a “very new product” that goes far beyond simply making driving safe and comfortable to making drivers unnecessary, it is difficult to just dismiss people’s fears that the cars will malfunction and cause accidents as being completely groundless. Recall the fear that robots will malfunction and injure or kill people in popular literary works like 2001: A Space Odyssey and I, Robot and the dramatic impression this fear has left on readers/viewers of these works. We cannot possibly say for sure that accidents due to the malfunction of robot cars would not have the same “dramatic effect” on people as these long-read/viewed classics.
On the one hand, suppliers and manufacturers first need to develop and implement engineering technology that sufficiently reduces the risk of malfunction-caused accidents before introducing it into the market. Such accidents can trigger a negative chain reaction ranging from condemnations of the product category as a whole to the branding of the manufacturer and other parties with a bad reputation. Those involved—especially the engineers—should, at the very least, avoid making hasty decisions as long as uncertainty remains. Given the magnitude of the repercussions of preventable malfunction-caused accidents, an error in judgment may lead to unending regret.
On the other hand, consumers and society as a whole need to calmly and objectively recognize the significant benefits offered by autonomous cars. Unfortunately, we cannot eliminate every risk posed by products—product accidents are inevitable. At the same time, however, autonomous cars can develop into a product category that offers substantial overall benefits as long as they do not cause accidents of abnormal frequency or severity. The introduction of autonomous cars into the market in this manner would eventually leave the world better off than it was before. Thus, it is imperative that consumers and society as a whole retain a calm, objective awareness of these benefits.