When autonomous vehicles cause accidents, the aggrieved parties will likely sue the car manufacturers, operators, or both for their injuries and damage to their property. (272) As control of the major operations of the vehicle transfers from the driver to the autonomous technology, it seems less likely that the traditional driver will be responsible for an accident caused by an autonomous vehicle in autonomous mode. (273) As I have argued elsewhere, and many others have, it is likely that tort liability will be imposed on the manufacturer of an autonomous vehicle for accidents caused by its autonomous technology. (274) This subsection will operate under the assumption that the manufacturer is liable.
This subsection analyzes the application of tort law to accidents involving the use of crash-optimization algorithms. (275) When a tortfeasor harms someone intentionally, the tortfeasor is typically sued for an intentional tort. (276) When the tortfeasor acts unintentionally but unreasonably, the remedy for the aggrieved party is in negligence. (277) Special issues arise when a person is injured by a product; when that occurs, the aggrieved party typically sues the manufacturer of the product under theories of products liability. (278) This subsection concludes by examining punitive damages--an overriding concern for all tortfeasors.
Because the decisions made by a crash-optimization algorithm are pre-programmed into the vehicle, those harmed by a crash optimization algorithm may believe they were "targeted." (279) For example, in the Motorcycle Problem, if the vehicle is programmed to hit the motorcyclist with a helmet, a helmeted motorist may believe that the vehicle "intentionally" hit her because of her helmet use. The person who believes she was targeted may seek recourse through an intentional tort action. (280)
The distinguishing feature of an intentional tort is intent on the part of the tortfeasor. A tortfeasor has the intent to commit an intentional tort if she (1) desires the consequences of the act or (2) is substantially certain that the consequences will result from the act. (281) The first method of satisfying intent is the common way people think about intent, and it clearly embodies wrongfulness. (282) An example would be when a tortfeasor aims her gun at and shoots a person. The second method of establishing intent is more convoluted. (283) A commonly used example of substantial certainty involves collateral harm (284): A sets up a bomb in B's office for the purpose of killing B, but A also knows that C is in the office and will die too, even though A does not intend to hurt C; nonetheless, A still detonates the bomb. (285) In that example, C's estate could sue A for the intentional tort of battery because A was substantially certain that C would die. (286)
A "crafty" plaintiff could assert that the algorithm writer commits the intentional tort of battery in the moral dilemmas from Part III. (287) For example, in the Trolley Problem, the estate of the person killed could assert that the car manufacturer programmed its vehicle with the intent to kill that person, or more likely, that the car manufacturer was substantially certain that its car would kill someone when doing so would result in more lives saved. Under the first method of establishing intent, it seems unlikely that a car manufacturer intended the harm; the programming of the vehicle and the actual accident is too attenuated in time and space to impute intent.
Under the substantial certainty test, imposing liability also seems too attenuated. (288) The comments to the Restatement (Third) of Torts state that the substantial certainty test is limited to:
[H]arm to a particular victim, or to someone within a small class of potential victims within a localized area. The test loses its persuasiveness when the identity of potential victims becomes vaguer and when, in a related way, the time frame involving the actor's conduct expands and the causal sequence connecting conduct and harm becomes more complex. (289) An accident caused by a crash-optimization algorithm seems "too broad and unfocused to support liability based on intent" (290)--the algorithm writer did not know who would be harmed or when such harm would occur. Moreover, the manufacturer programmed the crash-optimization algorithm to mitigate harm and damage, and certainly if the manufacturer had its choice, no one would ever be harmed by its car. (291) Therefore, because the identity of the individual victim of a decision made by a crash-optimization algorithm will not be known when the algorithm is programmed, and because the certainty of harm is "at some undefined time and place," (292) courts should not find the car manufacturer liable for an intentional tort. Instead, other tort doctrines better encompass injuries caused by the decisions of a crash-optimization algorithm.
Liability for most car accidents is analyzed based on principles of negligence, (293) which makes people liable for "unreasonably failing to prevent [a] risk." (294) To prove a negligence case, a plaintiff must show: (1) duty, (2) breach, (3) causation, and (4) damages. (295) A defendant must have a duty before she can be liable for injuries. (296) Generally, a defendant has a duty to exercise "reasonable care." (297) Another test used for determining the amount of care a defendant owed to a plaintiff is the "calculus of risk" test. (298) Judge Learned Hand developed a famous formula for the calculus of risk test:
[A defendant's duty of care] to provide against resulting injuries is a function of three variables: (1) The probability [of injury]; (2) the gravity of the resulting injury ... ; (3) the burden of adequate precautions. Possibly it serves to bring this notion into relief to state it in algebraic terms: if the probability be called P; the injury, L; and the burden, B; liability depends upon whether B is less than L multiplied by P: i.e., whether B less than PL. (299) A defendant breaches her duty when she fails to exercise reasonable care or when she engages in unreasonably risky conduct under the Hand formula. (300) Causation has two parts: cause-in-fact and proximate cause. (301) Generally, a defendant is a cause-in-fact of the harm to the plaintiff when that harm would not have occurred "but for" defendant breaching her duty. (302) Historically, courts have used two tests to determine whether the defendant's breach of her duty was the "proximate cause" of the plaintiffs harm: (1) directness and (2) foreseeability. (303) "The directness rule extends to all outcomes, even if not foreseeable, so long as they are the 'direct' result of the tortious conduct." (304) Under the foreseeability test, the defendant is only responsible for foreseeable results of her action--not freakish and unforeseen outcomes. (305) And finally, a plaintiff must prove that damages resulted from the defendant's negligence. (306)
These elements can be applied to an accident caused by a driver in a traditional vehicle. Assume that the driver was checking her cell phone and caused an accident when her vehicle crossed into oncoming traffic. In such a case, the driver has a duty to exercise reasonable care while driving. She breached that duty when she failed to pay attention to the road by checking her phone. The accident would not have occurred "but for" her using her phone; such an accident was foreseeable when she used her phone. Damages resulted from the accident. Thus, the driver is liable for the accident.
Negligence can also be applied to an accident that occurs because of a malfunction in the autonomous technology, and liability can be imposed on the car manufacturer for that harm. A car manufacturer has a "duty to exercise reasonable care to refrain from selling products that contain unreasonable risks of harm" to consumers and others who could be foreseeably harmed by its products. (307) It is also likely that an autonomous car will have a greater duty of care than a traditional driver because of its ability to discover danger better than a human driver. (308) A plaintiff cannot establish a breach by merely pointing out that a defect occurred--manufacturers only have to exercise reasonable but not perfect care. (309) Instead, the plaintiff must prove that the product was defective, and it was the manufacturer's negligence that caused the defectiveness. (310) The plaintiff would then need to prove that the car manufacturer's breach was the "but for" and proximate cause of her harm. (311) And then the plaintiff would need to establish damages. (312)
The principals of negligence can be applied to the moral dilemmas from Part III. In the Shopping Cart Problem, an operator has a duty to inspect her brakes, and a manufacturer has a duty to provide brakes that work. One breached that duty when the brakes failed. Had the operator inspected her brakes or the manufacturer supplied brakes in working fashion, no harm would have resulted--whether to the shopping cart, baby, or grocery store. That harm was a direct result of the breach, and it is foreseeable that failing to inspect brakes or supplying faulty brakes can result in a car accident. And damages did in fact occur. Therefore, the plaintiff could recover for negligence.
In the Bridge Problem, the driver of the school bus was negligent. That driver has a duty to drive on her side of the road. That driver breached her duty by driving in the wrong lane. If an accident ensues between the school bus and the vehicle, that breach will be the cause-in-fact and proximate cause of the accident, and damages would result. The question is whether the autonomous vehicle has a duty to avoid the accident. Generally speaking, a driver has a duty to exercise reasonable care to avoid an impending collision. (313) Failure to exercise reasonable care can be contributory or comparative negligence, which are affirmative defenses that...
Crashing into the unknown: an examination of crash-optimization algorithms through the two lanes of ethics and law.
|Author:||Gurney, Jeffrey K.|
|Position:||V. Legal Concerns B. Tort Law through VII. Conclusion, with footnotes, p. 224-267|
To continue readingFREE SIGN UP
COPYRIGHT TV Trade Media, Inc.
COPYRIGHT GALE, Cengage Learning. All rights reserved.
COPYRIGHT GALE, Cengage Learning. All rights reserved.