For Law Students

Join Now

Driverless car makers should not receive a liability free pass

Driverless Car
The Rinspeed Budii Concept autonomous car based on all-electric BMW i3 was presented at the 85th Geneva International Motor Show in 2015.

This year is the fifty-first anniversary of Ralph Nader’s Unsafe at Any Speed, which was the best-selling nonfiction book of 1966. The 1950’s and 1960’s were not happy days when it came to automotive safety.

Prior to the advent of products liability, the American automobile industry designed their vehicles for looks, not safety. Automobiles of the 1960s, even hard-top sedans, “crumpled like a Japanese lantern” in foreseeable rollover accidents.   In collisions, drivers were impaled by non-collapsible steering wheel columns and permanently tattooed by sharp edged emblems.   Because cars had no restraint systems, or even seat belts, occupants were frequently catapulted through windshields that were not made from laminated, safety glass. When someone was thrown from the car, through the windshield, they would be deeply lacerated by shards of glass. Car makers were not held legally accountable for the consequences of marketing dangerously defective automobiles because they routinely disclaimed warranties, limited remedies, and were shielded by the harsh doctrine of privity of contract.

The citadel of privity fell on May 9, 1960, when the Supreme Court of New Jersey announced the decision in Henningsen v. Bloomfield Motors, Inc.  In Henningsen, Chrysler sold an automobile to one of its dealers, Bloomfield Motors. The Chrysler dealer resold the new Plymouth to Claus Henningsen, whose wife, Helen, totaled the car only ten days after its purchase. The Plymouth’s steering gear malfunctioned, causing Mrs. Henningsen serious injuries when the out-of-control vehicle crashed into a wall. Chrysler contended that it was not liable because there was no privity and they had properly disclaimed all warranties except for replacing defective parts.  The New Jersey Supreme Court held the dealer and manufacturer liable, refusing to enforce Chrysler’s warranty disclaimers. The court reasoned that the lack of contractual privity was no longer relevant in the modern era of mass-marketed automobiles. Three years later, the California Supreme Court, adopted the reasoning of Henningsen, recognizing strict products liability for the first time. Today, almost every jurisdiction has adopted this progressive doctrine, permitting injured plaintiffs to recover for injuries caused by defective products of all kinds, without the necessity of proving manufacturer negligence.

Eighty years ago, in 1936, a Duke University law student published a note summarizing the status of automobile liability law. He observed that in 1905 all of American automobile case law could be contained within a four-page law review article, but three decades later, automobile liability law would “call for an encyclopedia.” That law student was Richard M. Nixon, who would later become President of the United States. His conclusion was that courts were mechanically extending “horse and buggy law” to automobiles. Just as in Nixon’s day, we are at a crossroads as to what liability rules should apply to the last incarnation of the “horseless carriage.” Commentators contend that we should not stretch products liability rules to driverless cars, as excessive products liability will choke off the potential benefits of this valuable technology. Industry advocates argue that we should instead turn to an alternative compensation system, modeled after the special fund for infants harmed by vaccines. Anyone injured by a driverless car could receive compensation from a fund established by auto makers, without the necessity of going to court.

We are highly skeptical that a no-fault compensation fund would work as effectively as products liability in creating manufacturers’ incentives for safety in driverless cars.  Between 1964 and 2013, products liability has helped to dramatically lower the number of fatalities per 100 million vehicle miles traveled from 5.39 to 1.10. Americans are safer because dangerously defective automobiles such as American Motor’s CJ-7 Jeeps, Ford’s rollover prone Explorers, and the notoriously flammable Ford Pintos have been redesigned or removed from the market after the manufacturer was found liable in products liability lawsuits. Vehicle makers may be tempted to short-change the public interest if products liability is foreclosed.  The risk of a manufacturer intentionally marketing defective vehicles has not disappeared. For example, Volkswagen, which announced its goal of being the industry leader in producing driverless cars, recently admitted to intentionally manipulating its software to fraudulently demonstrate that its diesel cars complied with EPA emissions standards.

The good news is that driverless cars have the potential to all but eliminate driver error, which should save large numbers of lives and greatly reduce accident insurance costs.  The annual death toll created by drunk, reckless, and distracted drivers will be eradicated.  However, driverless vehicles present novel safety, privacy, and security issues.  The German Federal Highway Research Institute concluded that the Tesla Model S’s autopilot feature constitutes a “considerable traffic hazard” because of both software and hardware defects.  Test vehicles made navigational errors while driving through a construction zone and where there were no road markings. The car’s sensors were not able to detect road conditions over 40 meters behind it, which makes it hazardous for driving on the autobahn, as there is no speed limit.

Highly vulnerable software controlling driverless cars may enable intruders to gain privileged access to computer systems, allowing intruders to alter code, misappropriate personal data, or destroy system files. Inadequately secured metadata gathered by connected cars may be misused by advertisers, law enforcement, insurers, and cybercriminals.   Security vulnerabilities in autonomous vehicles may enable attacks on the sensors that notify drivers of hazardous road conditions.

If the software controlling this vehicle is not sufficiently secured, a cybercriminal could override its settings to speed up the car or shift it into reverse, remotely endangering the car’s occupants or unwary pedestrians. Ransomware extortionists may use malware to disable cars’ safety devices or lock passengers into vehicles that have been immobilized in isolated settings. These risks are neither far-fetched legal hypotheticals nor science fiction fantasies. A Federal Trade Commission Workshop participant demonstrated that an attacker could gain “access to the car’s internal computer network without ever physically touching the car.” This terrifying scenario was explored last year in Season 7 of The Good Wife. In the episode entitled “Driven,” attorney Alicia Florick, played by Julianna Marguilies, is the story of how a hacker of a test driver of a vehicle disabled the driver’s ability to control the vehicle, causing a crippling accident.

The software industry should not be permitted to use contract law to shift the cost of defective code to the end user or consumer. Sun Microsystems’s Whitfield Diffie observed that many of the problems of computer security are due to bad software produced by firms that say, in effect, “‘You pay. We promise you nothing. Have fun.’ Just as in the contractual privity era, the software industry is using contract to escape accountability for dangerously defective code.  Consumers require strengthened tort remedies in the increasingly interconnected world of the Internet of Things.

Just as in Nixon’s day, products liability needs to continue evolving to provide consumers with remedies for malfunctioning vehicles and for those with inadequate security or privacy controls.  A strong products liability regime ensures that not even multibillion-dollar corporations can operate beyond the reach of the law. Tort law must continue to be forward-looking, with the flexibility to adapt to new technologies such as driverless cars. The public interest is best served by the natural evolution of well-established principles of tort law to driverless cars and other Internet-connected devices.

Acknowledgements: The authors would like to thank Professor Rustad’s research assistants Samantha Cannon, James Contrino, Maureen DeSimone, Samantha Lebrun, Robert Marin, and Priscilla Santos for valuable research and editorial suggestions. We also appreciate the useful editorial suggestions of Chryss J. Knowles.

Michael L. Rustad and Thomas Koenig Michael L. Rustad, Ph.D., J.D., L.L.M. is the Thomas F. Lambert Jr. Professor of Law and Co-Director of the Intellectual Property Law Concentration at Suffolk University Law School in Boston, Massachusetts. He is co-author with Thomas Koenig of "Global Information Technology: Ethics and the Law" (New York, New York: West Publishing Co., forthcoming 2017). Thomas Koenig has been a faculty member at Northeastern University since 1977, serving as chair of the Department of Sociology and Anthropology from 2002 until 2008. In 2008-2009 he was a Fulbright scholar at the University of Belgrade (Serbia) Law School, where he taught and researched European Internet policies.