BALANCING LIABILITIES IN AUTONOMOUS VEHICLE ACCIDENTS: DO NOT JUST BLAME THE GHOST INSIDE

By: Sean Kim*

I. Introduction

Despite its reemergence in the last five years, the concept of autonomous vehicles existed as early as the 1920s.[1]  While the early concept largely focused on speed and collision prevention systems provided by automated highway systems, or “smart roads,”[2] the advent of computers and artificial intelligence has shifted the focus from “smart roads” to “smart cars.”[3]  This shift has led to the continued development of vision-based systems of vehicle guidance.[4]  By 2007, all entrants in the United States Defense Advanced Research Projects Administration (DARPA) competition were successful in operating their autonomous vehicles in an urban setting that mimicked a city environment.[5]  As technology becomes more sophisticated, more automobile manufacturers and leading tech companies, such as Google and Apple, are cooperating to develop autonomous vehicles.[6]

As the number of autonomous vehicles on the road increases, ­a new genre of tort law is introduced: determining and balancing liability between drivers and manufacturers in autonomous vehicle accidents.[7]  While the general public expects autonomous vehicles to offer safer, hands-free driving experiences and to completely replace human interaction with motor vehicles, the need for consumer—or driver— education or behind-the-wheel training still exists.[8]

Section II provides an overview of negligence and strict liability under products liability law, as well as an overview of the types of automobile defects relevant to the analysis.  For the sake of this Note, breach of warranty will not be addressed.  Section III discusses recent developments of autonomous vehicle news and different state laws regulating autonomous vehicles.  Section IV recommends increased federal regulation for autonomous vehicles and heightened standards for vehicle manufacturers and drivers alike.

II. Background

A. Introduction to Products Liability

1. Negligence

Negligence is generally defined as “the failure to exercise reasonable care” under the circumstances.[9]  There are five elements required to establish a prima facie case for negligence: duty, breach of duty, “but-for” causation, proximate causation, and physical harm.[10]

Duty provides a maximum threshold to which people may be held accountable for their actions that cause harm to others.[11]  Without duty, one cannot be found liable for negligence.[12]  Breach of duty is often described as an “act or omission” that unreasonably affects the rights of others.[13]  While breach of duty implies a standard of reasonable care that people ought to follow to prevent undue harm to others,[14] the standard varies in different situations.[15]

The third and fourth elements of negligence are “but-for” and proximate causation.[16]  A causal relationship—both “but-for” and proximate—between a defendant’s breach of duty and the plaintiff’s harm must be established for liability to attach.[17]  “But-for” causation asks whether harm to the plaintiff would have happened were it not for defendant’s negligence,[18] while proximate causation relates to the closeness or remoteness of the defendant’s breach of duty to the plaintiff’s harm.[19]  The more remote a defendant’s action from a plaintiff’s harm, the less likely a court will find the defendant’s action a proximate cause of the plaintiff’s harm.[20]  The last element of negligence is actual harm.[21]  Without actual harm, no liability can be assigned to the defendant.[22]

2. Strict Liability

The doctrine of strict liability “[e]nsure[s] that the costs of injuries resulting from defective products are borne by the manufacturers that put such products on the market rather than by the injured persons who are powerless to protect themselves.”[23]  Strict liability attaches to a seller of defective or unreasonably dangerous products if the product causes harm to the user or consumer, or to his property, and if (a) the seller is engaged in the business of selling such a product, and (b) the product is expected to and does reach the consumer without substantial change in its condition.[24]  Although state courts apply different interpretations for the scope of strict liability,[25] almost all states have adopted the language of Section 402A of the Restatement in their rules.[26]

B. Types of Defects in Automobile Vehicles

1. Manufacture Defects

A manufacturing defect is a mistake in the process of building a product that would be safe if it were built as designed.[27]  A manufacturer may be held strictly liable for dangerous manufacturing defects, even if it has exercised “all possible care” in manufacturing the product.[28]  A plaintiff must establish that “the product does not conform to the specifications, regardless of whether there was negligence in the manufacturing process” to prevail in a products liability litigation.[29]  However, some courts hesitate to attach strict liability to software under the manufacturing defect doctrine.[30]

Another method available to consumers is the malfunction doctrine, a variation of the manufacturing defect doctrine.[31]  The malfunction doctrine allows a plaintiff to show a manufacturing defect by inferring product defect from “circumstantial evidence that (1) the product malfunctioned, (2) the malfunction occurred during proper use, and (3) the product had not been altered or misused in a manner that probably caused the malfunction.”[32]

2. Design Defects

A design defect is a defect in intended product design that makes a product harmful or dangerous.[33]  Many states assign strict liability to manufacturers for manufacturing design defects.[34]  The State of California uses two tests—the consumer expectation test and the risk/benefit test—to establish design defects­­[35] and to hold sellers and manufacturers strictly liable.[36]  Under the consumer expectation test, a plaintiff must prove that a defendant’s defective product did not perform as safely as an ordinary consumer would have expected it to perform when used or misused in an intended or reasonably foreseeable way, and that the product’s failure to perform safely was a substantial factor in causing the plaintiff’s harm.[37]  As for the risk/benefit test, strict liability attaches to a manufacturer when two conditions are satisfied: (1) the manufacturer’s product design was a substantial factor in causing harm to the plaintiff and (2) the manufacturer fails to prove that the benefits of the product’s challenged design outweigh the risks of such design.[38]

3. Failure to Warn

Manufactures may be held liable for injuries that are attributable to the relevant risks of their products.[39]  They are required to disclose and warn consumers of the foreseeable risks of using their products, and they can be liable for any injury or damage attributable to the lack of information disclosed.[40]  In California, for example, failure to warn (or inadequate warning) is a sufficient ground to hold manufacturers and sellers strictly liable.[41]  In order to minimize potential liability stemming from failure to warn, manufacturers err on the safe side by providing copious disclaimers and warnings.[42]

For example, Tesla’s new autonomous vehicle technology manual may have a disclaimer asking drivers to pay close attention to the traffic when using its autonomous vehicle technology.  Without a disclaimer warning drivers to pay close attention to the traffic while using the autonomous technology, Tesla could be held liable for any accident that may be attributable to this lack of disclaimer.[43]  In addition to providing adequate warning at the time of sale, manufacturers have post-sale responsibilities to provide warnings for newly discovered facts pertinent to the safety of their products.[44]  Although this post-sale responsibility is widely recognized by manufacturers, the “common law legal framework for addressing liability when manufacturers fail to do so is less well established.”[45]

III. Analysis

A. State Regulations for Autonomous Vehicle Drivers

As autonomous vehicle technology becomes more available to the public, a new standard may be required for drivers that utilize this technology.  Negligence is based on a standard that asks what a reasonably prudent person would do in a given, particular situation.[46]  In light of the perceived benefit of autonomous vehicle technology, however, the reasonable driver standard may shift, placing a higher burden on the vehicle and software manufacturers.

Basic reasonableness standards for autonomous vehicle drivers can be found in different state statutes.  For example, Nevada statute Sections 482A.070, outlining requirements for a human operator for highway testing of an autonomous vehicle,[47] and 482A.080, outlining equipment requirements for autonomous vehicles,[48] provide a basic requirement for autonomous vehicle drivers.  California, Florida, and the District of Columbia outline similar requirements for drivers and autonomous vehicles.[49]

These statutes generally include the following requirements: First, drivers must be seated in a position to allow them to immediately take control of the vehicle, i.e., being able to access the means to engage or disengage the technology.  Second, drivers must be able to safely monitor the autonomous vehicle operation.  Monitoring vehicle operation consists of monitoring the dashboard and hearing or seeing an alert indicating a failure of the autonomous system, or any other malfunction affecting the autonomous vehicle technology.  Third, drivers must be capable of taking control of the vehicle when necessary.  This requires that drivers not be in various physical or mental states—falling asleep behind the wheel or driving under the influence—that renders them incapable of manually driving the vehicle legally.

B. State Regulations for Autonomous Vehicles
1. Autonomous Vehicle Requirements

In Nevada, autonomous vehicles are required to provide a visual indication to its drivers when the autonomous technology has been engaged/disengaged, and be equipped with a means to alert the driver that the autonomous technology is unable to operate the vehicle safely.[50]  Moreover, Nevada statute requires that a driver of an autonomous vehicle be actively engaged while driving.[51]  For autonomous vehicle equipment and functionality, California, Nevada, Michigan, and the District of Columbia all require the vehicles to meet federal standards and regulations for motor vehicles and comply with applicable state traffic and motor vehicle laws.[52]  Furthermore, they require the vehicles to have safety mechanisms for engaging/disengaging the technology, visual indicators inside the vehicle that show when the vehicle is in autonomous mode, and a means of alerting the operator of a technology failure.[53]

2. Data Collecting and Reporting Requirements

One noticeable discrepancy between the aforementioned states is the requirement of reporting all disengagements of autonomous mode, near misses, and crashes.[54]  While California requires manufacturers to collect and report data related to accidents[55] or disengagement from autonomous mode by the test driver resulting from a failure of the autonomous technology,[56] Nevada only requires manufacturers to report accidents or traffic violations occurring during autonomous vehicle testing.[57]  Neither Florida nor the District of Columbia, on the other hand, has data collecting and reporting requirements for autonomous vehicle testing or operation.

3. Manufacturers’ Liability and Its Scope

For manufacturers, Nevada limits a manufacturer’s liability to accidents or injuries caused by defects that were present in the vehicle as originally manufactured.[58]  It thus shields manufacturers from potential liabilities from accidents or injuries caused by defects originating from any conversion or installation necessary to convert a regular vehicle into an autonomous vehicle.[59]  Michigan and the District of Columbia followed suit with Nevada’s approach.[60]  This approach to manufacturers’ liability seems fair and reasonable, and it closely follows the Restatement’s approach that if there were substantial changes in a product’s original condition, the manufacturer of the product would not be held liable.[61]

However, some jurisdictions have adopted a different approach.[62]  In 1996, an Illinois court wrote that “[w]here an unreasonably dangerous condition is caused by a modification to the product after it leaves the manufacturer’s control, the manufacturer is not liable unless the modification was reasonably foreseeable.”[63]  The “reasonably foreseeable” approach broadens the scope of potential liability of the automobile manufacturers because they could be held liable for damages or injuries caused by third-party modifications if these modifications were reasonably foreseeable to the manufacturer.[64]  On the other hand, California code does not even mention manufacturers’ liability, failing to address and balance liabilities between vehicle manufacturers and third­ parties that make modifications to manufactured vehicles.[65]

C. Determining a Standard for Manufacturers Liability

Further development and public use of autonomous vehicle technology may challenge the adequacy of current product liability law.  It has been suggested that the advent of autonomous vehicles will result in an imbalance of liability between autonomous vehicle manufacturers and consumers.[66]  After a much-publicized fatal accident that involved a driver with his Tesla Model S electric sedan in autonomous driving mode, federal regulators opened a formal investigation into the accident.[67]  Although Tesla, with possible recall of its vehicles looming (depending on U.S. National Highway Traffic Safety Administration (NHTSA) findings), escaped from the aforementioned accident unscathed, the accident nevertheless highlighted the need for manufacturers to eliminate any software and/or hardware defects and to provide all required information to drivers and vehicle owners.[68]

The Restatement (Third) of Torts recognized that the seller’s duty to warn of product-related defects after the point of sale is “often daunting.”[69]  Continued technological developments, such as on-board sensors and driver assistance systems—adaptive cruise control, automated emergency braking, and pedestrian detection[70]—make automakers more vulnerable to negligence and strict liability.[71]  This increase in liability, so-called “proximity-driven liability,” resulting from increased proximity between drivers and automobiles, may run against the spirit of the doctrine of strict liability, which is to safeguard the general public from defective products by increasing liability of the manufacturer—the only party able to rectify the defect and prevent public loss.[72]

The purpose of autonomous vehicle technology is to make the vehicle safe for consumers.[73]  Knowing that software is never perfect, however, it is unfair to place a higher burden of liability on manufacturers that aim to make driving easier and safer for the general public.  Barring specific instances in which software failure or malfunction is the sole cause of an accident, autonomous vehicle operators should also shoulder the responsibility of being safe operators.  Nevertheless, vehicle manufacturers should shoulder any liability stemming from defects of its hardware and/or software, regardless of whether it was negligent.

IV. Recommendation

A. Standards for Manufacturers and Drivers

Courts have generally refused to apply the doctrine of strict liability to software failures because of the notion that software cannot be perfect and error-free.[74]  While understandable, strict liability should be applied to lessen the burden borne by the public—the burden of potential dangers and both the economic and social costs associated with automotive accidents.

Manufacturers and developers are in a far better position to prevent and mitigate software failures or defects.  Furthermore, increasingly sophisticated marketing of autonomous vehicles and the autonomous vehicle technology makes it more difficult for the consumers to see the risk behind the technology.[75]  Moreover, complicated chains of supply of parts and distribution of autonomous vehicles make it that much more difficult for consumers to pinpoint the origin of the defect in a manufacturer’s product.[76]  Some autonomous vehicle manufacturers such as Google and Mercedes-Benz have indicated that they will take full responsibility of any accidents caused by failures of their autonomous vehicle software.[77]

Moreover, courts should apply the objective reasonable-person standard to autonomous vehicle software, which should consider the following, non-exhaustive factors in determining whether a software was defective: (1) total utility, or benefit, to the drivers and others, (2) total amount of risk, or harm, to the driver and others, (3) the likelihood of the risk actually causing harm to others, and (4) any existing, reasonable alternatives of lesser risk and the costs of those alternatives.[78]  This standard should allow courts to determine whether the software made a reasonable—as opposed to correct—decision, eliminating potential ethical dilemmas from the liability calculus. [79]  Although the reasonable-person standard lacks certainty,[80] courts, nevertheless, have been successfully applying it to many tort cases, and thus should be able to determine whether autonomous vehicle software has acted reasonably according to the standard.

As autonomous vehicle technology becomes more advanced and ready for public use, autonomous vehicle drivers may become less responsible on the road, creating a potential imbalance of liability between manufacturers and consumers.[81]  Courts should apply a higher standard to autonomous vehicle drivers involved in an automobile accident while using autonomous vehicle technology.  This will require the courts to determine whether drivers of autonomous vehicles were driving, or monitoring their autonomous vehicles in a reasonable way.  Despite possible concerns about uncertainty and the unpredictable nature of the reasonable-person standard, its application would not be difficult for the courts since they have been more than capable of determining whether one has acted reasonably or not in a given situation.

B. Federal Regulation for Autonomous Vehicle Technology

Congress should enact a law regulating autonomous vehicle operations, especially to require sensors and software functionality, including parameters that govern and influence autonomous vehicle software’s decision making.  The United States Department of Transportation (USDOT) already has extensive safety standards and regulations provided by NHTSA for regular vehicles.[82]  As evidenced in numerous state laws, bills, and regulations on autonomous vehicles, regulations and requirements for the operation of autonomous vehicles are scant and general at best.[83]  On the other hand, for aircrafts, the Code of Federal Regulations (CFR) and the Federal Aviation Administration (FAA) have extensive requirements and regulations for flight guidance systems.[84]

An example of a possible federal regulation on autonomous vehicles may be monitoring and recording autonomous vehicle and driver activities for a specified time period before an accident.  The recorded vehicle and driver activities would greatly help courts determine whether the driver was monitoring his or her autonomous vehicle operation in a reasonable manner.  Furthermore, the government should regulate technical parameters such as sensor sensitivity, radar range, and software processing speed, and delineate what constitutes a substantial change, or material modification, to autonomous vehicles and autonomous vehicle software.  Unified quality standards for autonomous vehicle parts and sensors would provide clarity and information to consumers and greater control and guidance on autonomous vehicle performance, safety, installation, modification, and testing standards.

V. Conclusion

Autonomous vehicle technology is no longer a product of science fiction.  Google’s self-driving vehicles have driven over two million miles so far,[85] and many Tesla drivers are already taking advantage of Tesla’s autonomous vehicle software.[86]  However, the technology is far from perfect, and there are scenarios that are beyond technology’s current capabilities to handle.[87]  In order to protect the public from accidents caused by autonomous vehicle software defects or malfunctions, standards for autonomous vehicle manufacturers should be heightened.  Likewise, in order to protect the public from accidents caused by negligent drivers using this technology, reasonableness standards for these drivers should likewise be heightened.[88]

In addition, a federal department such as the Department of Transportation, or an agency such as NHTSA, should regulate autonomous vehicle operation, including sensor and radar operations, software controls and parameters, vehicle inspection guidelines for manufacturers and third parties, and operation manuals for autonomous vehicle drivers.  Centralized federal regulation would provide concrete guidelines not only for the states but also for the vehicle manufacturers and the public alike.  The technology is already vastly ahead of the regulation, and there is some serious work to do for our state and federal legislatures.


* Juris Doctor, University of Illinois College of Law, 2017. Thanks to the editors and staff of the Journal of Law, Technology & Policy for their efforts. I would also like to thank my wife, Catherine, for her continuous support. This work would not have been possible without her. Finally, many thanks are owed to my parents, family, and friends for their unwavering support.

[1] Marc Weber, Where to? A History of Autonomous Vehicles, Computer History Museum, http://www.computerhistory.org/atchm/where-to-a-history-of-autonomous-vehicles (last visited Apr. 8, 2017).

[2] Id.  This was because much of the danger from driving during that time period was from ill-marked roads rather than the automobiles themselves.

[3] Id.

[4] Id.

[5] Id.

[6] 33 Corporations Working on Autonomous Vehicles, CB Insights (Aug. 11, 2016), https://www.cbinsights.com/blog/autonomous-driverless-vehicles-corporations-list.

[7] John Villasenor, Products Liability and Driverless Cars: Issues and Guiding Principles for Legislation, Brookings (Apr. 24, 2014), http://www.brookings.edu/research/papers/2014/04/products-liability-driverless-cars-villasenor.

[8] Sherry Baxter, Reasonable Doubt: The Road to Regulation for Self-Driving Vehicles, Ga. Straight (Jan. 22, 2016, 2:17 PM), http://www.straight.com/news/622601/reasonable-doubt-road-regulation-self-driving-vehicles.

[9] See, e.g., Bodin v. City of Stanwood, 927 P.2d 240, 249 (Wash. 1996) (stating basis for negligence action).

[10] David G. Owen, The Five Elements of Negligence, 35 Hofstra L. Rev. 1671, 1674 (2007).

[11] Id. at 1675.

[12] Palsgraf v. Long Island R. Co., 162 N.E. 99, 99 (N.Y. 1928).

[13] Id.

[14] Id.

[15] While adults are held to a reasonable person standard, children and disabled people are held to a standard of reasonableness for a person with similar characteristics.  Restatement (Third) of Torts § 10 (2010); see also Stevens v. Veenstra, 573 N.W.2d 341 (Mich. Ct. App. 1997) (holding that a fourteen-year-old driver education student is not held to a reasonable person standard for adults, but instead to a standard reasonable for fellow fourteen-year-olds).  However, people with greater levels of skills, such as doctors and other medical professionals, are required to exercise a greater amount of care they reasonably possess.  See Helling v. Carey, 519 P.2d 981, 983 (Wash. 1974) (finding that defendant was negligent in failing to conduct a simple pressure test that other optometrists would reasonably have done).

[16] Id.

[17] Id.

[18] Id.

[19] Id.

[20] Id.

[21] Id.

[22] Id.

[23] Greenman v. Yuba Power Products, Inc., 377 P.2d 897, 901 (Cal. 1963).

[24] Restatement (Second) of Torts § 402A (1965).

[25] Villasenor, supra note 7.

[26] Derek H. Swanson & Lin Wei, McGuireWoods, United States Automotive Products Liability Law (Oct. 2009), https://www.mcguirewoods.com/news-resources/publications/us-automotive-products-liability.pdf.

[27] Restatement (Third) of Torts: Prod. Liab. § 2(a) (1998).

[28] Id.

[29] Jeffrey K. Gurney, Sue My Car Not Me: Products Liability and Accidents Involving Autonomous Vehicles, 2013 U. Ill. J.L. Tech. & Pol’y 247, 258 (2013).

[30] See 68 Am. Jur. 3d Proof of Facts § 8, at 333 (2002) (“[N]o cases have been found applying [manufacturing defects] to software.”).

[31] David G. Owen, Manufacturing Defects, 53 S.C. L. Rev. 851, 873 (2002).

[32] Id.

[33] Restatement (Third) of Torts: Prod. Liab. § 2(b) (1998).

[34]Dennis W. Stearns, An Introduction to Product Liability Law, Marler Clark L.L.P., http://www.marlerclark.com/pdfs/intro-product-liability-law.pdf (last visited Apr. 8, 2017).

[35] Barker v. Lull Eng’g Co., 573 P.2d 443, 457–58 (Cal. 1978).  The California Supreme Court in Barker set out two tests to establish defect in the product design: the consumer expectation test and the risk/benefit test.  Id.  For the consumer expectation test, a plaintiff must prove that the defendant’s defective product did not perform as safely as an ordinary consumer would have expected it to perform when used or misused in an intended or reasonably foreseeable way, and that the product’s failure to perform safely was a substantial factor in causing plaintiff’s harm.  Id.  The consumer expectation test is reserved for cases where plaintiff’s everyday experience permits a conclusion that the product design is defective and not safe.  Pruitt v. General Motors Corp. 72 Cal. App. 4th 1480, 1484 (1999).  For the risk/benefit test, the plaintiff has to prove that the defendant’s product design was a substantial factor in causing harm to the plaintiff, and the defendant must fail to prove that the benefits of the product’s challenged design outweigh the risks of the design.  Barker, 573 P.2d at 457–58.

[36] David H. Canter et al., California Products Liability Law: A Primer (Jan. 2012), http://trialattorneysofamerica.com/documents/Primer2012.pdf; see also Anderson v. Owens-Corning Fiberglas Corp., 810 P.2d 549, 553 (Cal. 1991) (holding that strict liability has been invoked for three different types of defects: manufacturing, design, and inadequate warnings).

[37] Barker, 573 P.2d at 457–58.

[38] Id.

[39] Restatements (Third) of Torts § 2(c) (1998); see also Villasenor, supra note 7.

[40] Villasenor, supra note 7.

[41] In Livingston v. Marie Callenders, Inc., the court found Marie Callenders liable to a plaintiff who suffered an allergic reaction to a Marie Callenders product on a strict liability failure to warn theory.  72 Cal. App. 4th 830 (1999).  The product contained an ingredient (MSG) to which a substantial number of the population is allergic.  Id. at 832–33.  Also, the ingredient was one whose danger was not generally known, or if known was one which the consumer would reasonably not expect to find in the product, and where the defendant knew, or by the application of reasonable developed human skill and foresight should have known, of the presence of the ingredient and the danger.  Id.

[42] Villasenor, supra note 7.

[43] A product is defective because of inadequate warnings if the foreseeable risks of harm posed by the product could have been reduced by reasonable warnings by the seller or other distributor, and the lack of warnings renders the product not reasonably safe.  Restatement (Third) of Torts: Prod. Liab. § 2(c) (1998).

[44] Id. § 10.  The Supreme Court of Michigan ruled in Comstock v. General Motors Corp. that a manufacturer of an automobile (GM), the brakes of which were defective, had a duty to warn of the vehicle’s inherent danger, not only at the time of sale but any time shortly after the product entered the stream of commerce if a defect became known to the manufacturer and if a failure to give prompt warning of the known, latent defect imperiled life and limb.  99 N.W.2d 627, 636 (Mich. 1959).  In Hasson v. Ford Motor Co., the Supreme Court of California stated in its dicta that the manufacturer was negligent in failing to take into account (in designing its braking systems and advising on their maintenance) foreseeable brake abuse by drivers resulting in overheating of brakes.  564 P.2d 857, 870 (Cal. 1977).

[45] Villasenor, supra note 7.

[46] Owen, supra note 10, at 1677.

[47] Nev. Rev. Stat. § 482A.070 (2013).

[48] Id. § 482A.080.

[49] See Cal. Code Regs. tit. 13, § 227.44 (2014); Fla. Stat. §319.145 (2012); D.C. Code §50-2352 (2012) (outlining some vehicle and driver requirements for testing and/or operating autonomous vehicles).

[50] Nev. Rev. Stat. § 482A.080 (2013).

[51] Id. § 482A.070.  Autonomous vehicle drivers are required to be (a) seated in a position so that he can take immediate control of his vehicle, (b) able to monitor safe operation of the vehicle, and (c) capable of taking control over the autonomous vehicle in case of emergency.

[52] James M. Anderson et al., RAND Corp., Autonomous Vehicle Technology: A Guide for Policymakers 44–47 (2016), http://www.rand.org/content/dam/rand/pubs/research_reports/RR400/RR443-1/RAND_RR443-1.pdf.

[53] Nev. Rev. Stat. § 482A.080 (2013); D.C. Code §50-2352 (2012); Cal. Veh. Code § 38750(c) (2013); Fla. Stat. §319.145 (2012).

[54] Technology Law & Policy Clinic Autonomous Vehicles Team, Autonomous Vehicle Law Report and Recommendations to the ULC, U. Wash. Sch. L., https://www.law.washington.edu/Clinics/Technology/Reports/AutonomousVehicle.pdf, at 5 (last visited Apr. 8, 2017).

[55] Cal. Code Regs. tit. 13, § 227.44 (2014).

[56] Id. § 227.46.

[57] Nev. Admin. Code §482A.130 (2012).

[58] Id. § 482A.090.

[59] Id.

[60] “A manufacturer of automated technology is immune from civil liability for damages that arise out of any modification made by another person to a motor vehicle or an automated motor vehicle, or to any automated technology . . . .”  Mich. Admin. Code r. 257.817 (2013).  “The original manufacturer of a vehicle converted by a third party into an autonomous vehicle shall not be liable in any action resulting from a vehicle defect caused by the conversion of the vehicle, or by equipment installed by the converter, unless the alleged defect was present in the vehicle as originally manufactured.”  D.C. Code §50-2353 (2013).

[61] “One who sells any product in a defective condition unreasonably dangerous to the user or consumer or to his property is subject to liability for physical harm thereby caused to the ultimate user or consumer, or to his property, if (a) the seller is engaged in the business of selling such a product, and (b) it is expected to and does reach the user or consumer without substantial change in the condition in which it is sold.”  Restatement (Second) of Torts § 402A(1) (1965) (emphasis added).  However, a question remains: what constitutes a substantial change?  The Restatement leaves that question to the courts to decide.  See id. § 402A(1) cmt. p (commenting on varying degree of changes that can be made to a product without shifting the liability from the seller or manufacturer of the product to the party that made the changes).

[62] Villasenor, supra note 7.

[63] Davis v. Pak-Mor Mfg. Co. 672 N.E.2d 771, 775 (Ill. Ct. App. 1996); see also Woods v. Graham Eng’g Corp., 539 N.E.2d 316, 318  (Ill. Ct. App. 1989) (ruling that when a change or modification by a third party is foreseeable, liability will still be imposed on the original manufacturer); Hoyt v. Wood/Chuck Chipper Corp., 651 So.2d 1344, 1351 (La. Ct. App. 1995) (holding engine manufacturer not liable since it could not have anticipated plaintiff’s material alteration of the power unit).

[64] See Lavoie v. Power Auto, Inc., 312 P.3d 601, 608 (Or. Ct. App. 2013) (holding that changing a floor mat of a vehicle was a reasonably foreseeable modification/alteration that was a substantial contributing factor to the personal injury).  So in some states, automobile manufacturers will have to consider changing floor mats as one of the foreseeable “substantial” modifications that can cause automobile accidents.

[65] Cal. Veh. Code § 38750 (2015); see also Anderson et al., supra note 52, at 47.

[66] Gary E. Marchant & Rachel A. Lindor, The Coming Collision Between Autonomous Vehicles and the Liability System, 52 Santa Clara L. Rev. 1321, 1339 (2012).

[67] Bill Vlasic & Neil E. Boudette, Self-Driving Tesla Was Involved in Fatal Crash, U.S. Says, N.Y. Times (June 30, 2016), https://www.nytimes.com/2016/07/01/business/self-driving-tesla-fatal-crash-investigation.html.

[68] Id.

[69] Restatement (Third) of Torts § 10 cmt. a (1998).

[70] Sven A. Beiker, Legal Aspects of Autonomous Driving, 52 Santa Clara L. Rev. 1145, 1147–48 (2012).

[71] Bryant Walker Smith, Proximity-Driven Liability, 102 Geo. L.J. 1777, 1794 (2014).

[72] Id. at 1778.

[73] FAQ, Waymo, https://waymo.com/faq/ (last visited Apr. 8, 2017).

[74] 68 Am. Jur. 3d Proof of Facts § 8, at 333 (2002).

[75] Frances E. Zollers et al., No More Soft Landings for Software: Liability for Defects in an Industry that Has Come of Age, 21 Santa Clara Computer & High Tech. L.J. 745, 746 (2005).

[76] Id.

[77] Michael Ballaban, Mercedes, Google, Volvo to Accept Liability when Their Autonomous Cars Screw Up, Jalopnik (Oct. 7, 2015, 11:47 AM), http://jalopnik.com/mercedes-google-volvo-to-accept-liability-when-their-1735170893.

[78] The factors closely mimic that of the reasonable-person standard that reflects a cost-benefit approach supported by principles of utility and efficiency.  Owen, supra note 10, at 1677.  It also resembles the “Hand Formula” that uses a risk-calculating approach of judging one’s decision making.  United States v. Carroll Towing Co., 159 F.2d 169, 173 (2d Cir. 1947).

[79] John Gardner, The Many Faces of the Reasonable Person, N.Y.U. Sch. L., at 7­–8, http://www.law.nyu.edu/sites/default/files/upload_documents/The%20Many%20Faces%20of%20the%20Reasonable%20Person.pdf.

[80] Id.

[81] See Alex Davies, Obviously Drivers Are Already Abusing Tesla’s Autopilot, WIRED (Oct. 22, 2015, 7:00 AM), http://www.wired.com/2015/10/obviously-drivers-are-already-abusing-teslas-autopilot/ (noting that some drivers of autonomous vehicles are driving recklessly by pushing the limit of the autonomous vehicle technology).

[82] Federal Motor Vehicle Safety Standards and Regulations, Nat’l Highway Traffic Safety Admin., http://www.nhtsa.gov/cars/rules/import/FMVSS (last visited Apr. 8, 2017).

[83] The Hawaii House of Representatives’ bill for autonomous vehicle operation only includes safety requirements, minimum insurance coverage, and necessary equipment and performance standards that ensure safe operation.  H.R.B. No. 1458, 28th Leg. (Haw. 2015).  In addition, Nevada’s statute has not set forth autonomous vehicle requirements for operations, insurance, and minimum safety standards.  Nev. Rev. Stat. § 482A.100 (2013).

[84] 14 C.F.R. § 25.1329 (2016).

[85] Waymo, supra note 73.

[86] Your Autopilot Has Arrived, Tesla (Oct. 14, 2015), https://www.teslamotors.com/blog/your-autopilot-has-arrived.

[87] Neal E. Boudette, Tesla’s Self-Driving System Cleared in Deadly Crash, N.Y. Times (Jan. 19, 2017), https://www.nytimes.com/2017/01/19/business/tesla-model-s-autopilot-fatal-crash.html.

[88] See Sam Levin & Nicky Woolf, Tesla Driver Killed While Using Autopilot Was Watching Harry Potter, Witness Says, Guardian (July 1, 2016, 1:43 PM), https://www.theguardian.com/technology/2016/jul/01/tesla-driver-killed-autopilot-self-driving-car-harry-potter (noting that driver negligence may have contributed to the first fatal Tesla crash).