The GULS Law Review

Getting you through the GU law degree!

header photo

Self-Driving Cars: the legal ‘whodunit’

February 17, 2017

In this article, 4th Year LLB student and Obligations Law sub-editor Sarah Hassall,  addresses whether or not the present law is suitable for the incorporation of autonomous vehicles. It looks to the developing law in the USA concerning issues of liability and considers how these issues would be handled in Scotland.

Self-Driving Cars: the legal ‘whodunit’

We are living in the 21st century, an age where technology is taking off and achieving incredible things far beyond what we could imagine. The latest invention to (quite literally) hit the streets is the concept of cars without drivers. Google has been developing and testing these autonomous vehicles in the USA, and companies like Ford and Tesla have also begun their own developments[1]. While preliminary issues involving adjusting legislation to accommodate these new vehicles are being resolved in several states, a new wave of problems are beginning to emerge. After a series of accidents there are claims that the drivers ought to have been paying attention, however in a car which requires no human input this seems unjustifiable. This article will discuss this duty and whether or not it is necessary. It will also address product liability, the main avenue for suit in the USA. I intend to look at these issues in the States and discuss how they would be handled in Scotland.


The Driver’s Duty

 With the exception of cases involving contributory negligence, road traffic accidents are relatively straightforward and the wrongdoer easily identifiable. In ordinary practice the driver of the vehicle in the wrong is at fault and a case will be brought against him for the damages, with his losses being indemnified under his insurance policy. We are now, however, presented with a case in which there is no driver. There is the manufacturer, the system, and the ‘operator’ of the vehicle. The vision manufacturers and consumers alike have for autonomous vehicles is that we step into our car in the morning, push a button, and relax while our car drives us off to work.[2] However development is a process and we must ask ourselves the age-old question: are we there yet? Manufacturers and legislators don’t think so but the general public is sold. This leads to the first question to be answered: is there a duty on the driver to remain alert while operating the vehicle?

            In the wake of several accidents in the US, Tesla has made several statements about the operation of their vehicles. After an unfortunate death, the company stated that “the driver is still responsible for, and ultimately in control of, the car.”[3] This is not just a wild claim made by the company to avoid liability either; regulations in California echo this sentiment. §227.86 of the draft Regulations concerning autonomous vehicles states that “in the event that the system requires the driver to take control of the vehicle or when the vehicle is operating outside of its approved operational design domain, the driver shall be responsible for the safe operation of the vehicle.”[4] This is clear acknowledgement that the technology is not perfect and may require a manual override. Although this may be the case, Douma and Palodichuk raise the question of how far this duty extends. They note “the issue is whether the operator has a duty to continually monitor the behavior [sic] of the car, being ready to take over at any moment, or if the operator is only obligated to respond to safety alerts generated by the vehicle.”[5] Tesla have been firm in their answer. The company has stated that the system is supposed to remind drivers to keep their hands on the wheel[6] and that not having his or her hands on the wheel is contrary to the system’s terms of use.[7]

            It has been acknowledged that should the car be functioning in autonomous mode the manufacturer will be responsible for any and all traffic violations.[8] This statement relieves concerns of the burden weighing too heavily on the operator for errors, as a previous version of the regulations reversed the liability.[9] It has been well documented that the vehicles are not flawless; when discussing tests run by Google, Cohen reported that “the vehicle was reported to have difficulty identifying when objects, such as trash and light debris, are harmless, causing the vehicle to veer unnecessarily,”[10] so the revised edition provides much needed recognition that system errors cannot be attributed to the operator. This also introduces a degree of flexibility to the regulations to accommodate for these scenarios, especially when the technology is a work in progress.

It has been suggested by several American commentators that the standards for autonomous vehicles ought to be similar to those of aviation. Again, Douma and Palodichuk open the discussion by asking “is the ‘driver’ of an autonomous vehicle like the engineer of a train or pilot of an aircraft on ‘autopilot’, or is she simply a passenger?”[11] The logic behind this connection is that pilots are required to observe and maintain control over the autopilot technology, so in the event of a failure they can immediately override the system. However, the problem with this association is that the autopilot technology in planes “is not designed to adjust to a changing environment or adjust the course of the vehicles without input from a human controller,”[12] unlike an autonomous vehicle is designed to do. There is a fundamental difference in purpose between the autopilot system in a plane and those in autonomous vehicles, so it is unlikely that the application of aviation standards to operators of these vehicles would withstand challenge.

            In the UK, however, there are no specific rules dedicated to autonomous vehicles presently. This would leave them to be governed by the Road Traffic Act 1988 (RTA). There are references throughout the Act to a ‘competent and careful driver,’[13] confirming that drivers do indeed have a general duty to operate the vehicle in a reasonable manner. Furthermore, there is specific provision for requiring drivers to maintain ‘proper control’ of the vehicle.[14] It is apparent, particularly with reference to the control requirements, that this legislation is not satisfactory for autonomous vehicles since their operational requirements are fundamentally different from current road-legal cars. Nevertheless, it does not detract from the necessity and applicability of the underlying ‘reasonableness’ principle. This can be observed at the common law where the driver of an ambulance in a collision was not found negligent because she had done all that she could reasonably do in the circumstances.[15] The standard for reasonableness in autonomous vehicles is not yet clear; whether the end result is a legal requirement that the driver remain alert and able to manually control the vehicle if necessary remains to be seen. What is clear from the California regulations is that some degree of alertness ought to be maintained while operating autonomous vehicles.


Balancing the Duty with Expectations

Although the Californian regulations provide for a duty to remain alert, is this duty actually necessary? Is it not the case that the driver, as a consumer, should be entitled to a reasonable expectation that the system will not fail? This question can only be raised because of the circumstances under which these cars are being marketed. While it is entirely unreasonable to expect any system to be flawless, the primary selling point of the autonomous vehicles is their safety and complete control with no input from the driver. Consumers are certainly of the opinion that once autopilot is engaged, there is no longer any requirement for them to remain involved. However, as is evidenced by accidents, system failures do happen and may require operator input. In one case a man was watching a movie in his car when the system failed to distinguish a trailer crossing the road, resulting in a crash and the untimely death of the driver.[16] While this example is interesting for several reasons, it demonstrates the need for the operator to occasionally control the vehicle.

            What level of performance are consumers entitled to expect? In the US, “a consumer must have sufficient knowledge of a product’s design to develop a reasonable expectation about its performance or safety.”[17] The wording of this requirement is concerning. In his discussion of this standard, Herd suggests that the complexity of the technology may restrict the application of this standard.[18] When assessed pedantically, the average consumer will not have ‘sufficient knowledge’ of the way in which the system operates and controls the vehicle, so arguably the consumer cannot reasonably expect a certain degree of functionality. Additionally, there are concerns in America that “consumers will develop unrealistic expectations about how well the cars can maneuver [sic] and avoid accidents, causing some consumers to become over-reliant on the technology.”[19] While this is a perfectly valid concern, it should not eliminate the possibility that a consumer has developed some expectations for the car’s performance. After all, one cannot market the vehicle on the premise that it can drive autonomously and not allow for an expectation that the system will do so the majority of the time.

            In the UK, the Consumer Protection Act 1987 (CPA) s.3 sets out circumstances to take into account when considering what a person is entitled to expect. This includes “the manner in which, and purposes for which, the product has been marketed,”[20] and “what might reasonably be expected to be done with or in relation to the product.”[21] This appears to be a wider test than the American one described above, with greater appreciation for the way in which the product has been marketed and less focus on an understanding of the system. And while there is a certain expectation from consumers and producers that, as regards safety, autonomous vehicles will outperform ordinary drivers, a system failure does not automatically indicate a defect. Determining the standard of reasonableness will be challenging; it is evident from public reactions to accidents that an incredibly high burden has been placed on these vehicles, and while no system can be flawless there is some expectation that they should be.

Product Liability Problems

It is widely agreed that product liability will be the appropriate avenue for action when accidents do occur. Marchant and Lindor describe the liability requirements in the US as having three categories of defect, one of which must be found in order to proceed with a claim. The categories are: 1) a manufacturing defect; 2) failure to provide adequate warnings; and 3) a design defect.[22] Of these, the design defect requirement, otherwise known as the risk-utility test, is the centre of most discussions. This test “requires the plaintiff to demonstrate that his injuries were proximately caused by the product design and that the injuries could have been prevented by a reasonable design alternative.”[23] This places an extreme amount of pressure on the pursuer to demonstrate this connection and suggest a superior alternative. It has also been noted that this test could “make it difficult to find qualified experts with legitimate experience, and simply make it too expensive to pursue claims.”[24] Furthermore, because we are only at the beginning of a long process of development, it is unlikely at this time that there would be a ‘reasonable design alternative’ other than an ordinary car, so claims pursued under this avenue fall flat before they have properly begun.

            Frankly, the common law position in the UK is not much better. While strict liability has been introduced by the CPA, not all product liability cases will meet the requirements and so will fall to the case law. The original formulation of this test comes from Donoghue v Stevenson,[25] where the question asked, and answered in the affirmative, was “whether the manufacturer … in circumstances which prevent the distributor or the ultimate purchaser or consumer from discovering by inspection any defect, is under any legal duty to the ultimate purchaser or consumer to take reasonable care that the article is free from any defect likely to cause injury to health.”[26] This case established the concept of the duty of care and forms the basis of the entirety of modern negligence. Subsequent case law, however, has been particularly unhelpful. In the same year, it was simultaneously held that in cases of product liability the onus is on the pursuer to point to the origin of the defect,[27] and that the exact location of the defect is not necessary for a claim in negligence,[28] with neither case referencing the other. This lack of clarity potentially leaves us in much the same predicament as in the USA. With a complex system created from multiple products and by several manufacturers, it becomes almost impossible to attribute an accident to one specific part. Strict liability removes the burden of locating the defect as long as the test’s components are met. The CPA establishes that if damage is caused by a defect, the producer, or anyone holding himself to be the producer, will be strictly liable for the damage.[29] The consumer expectations test discussed above assists in defining what a defect is; arguably this part of the test will provide the most trouble for autonomous vehicles for the reasons previously discussed.

Case law following the CPA offers guidance as to what will be considered a defect. There are two cases in particular that, I believe, offer substantial difficulties for claiming defects in autonomous vehicles. The first is the case of Richardson v LRC Products,[30] which involved a defective condom leading to an unwanted pregnancy. It was held that a product is not defective if the public know there is a risk that the product might fail in its purpose. It seems trite to mention that drivers are well aware of the risk of accidents when operating any vehicle, and given that it is impossible for technology to run flawlessly it is feasible that similar arguments can be made successfully against autonomous vehicles.[31] The second case is B (a child) v McDonald’s Restaurants ltd,[32] where it was held that a product is not defective if the risk was obvious. This case, perhaps infamously, involved arguments that coffee cups were not adequate for serving hot drinks, and there was not sufficient warning that the contents were hot. These arguments were not successful because the risk of scalding from hot coffee was so obvious that a failure to warn or use an alternative cup was not necessary. Justice Field also stated that “it was up to those frequenting the restaurants to take care not to drop or knock over hot drinks.”[33] His comments suggest an underlying duty to avoid obvious risks where possible and within reason. Again, what can be taken from this case is that car accidents could be deemed such an obvious risk that the presence of an accident does not render a car defective. This is an extremely valid point, however its applicability to the autonomous system may not be so straightforward. It would be reasonable to argue that system failure is not an ‘obvious risk,’ but this argument could also be countered by suggesting that a driver remaining alert is a means of taking reasonable steps to avoid the obvious risk from manifesting.[34] The CPA seems to be much better equipped to handle the problem of autonomous vehicles; certainly, it removes the onus from Evans that would have proved itself quite problematic. Nevertheless, key issues ought to be addressed before the legislation is fully capable of addressing autonomous vehicles.



In this brief assessment of the legal climate in the USA as regards autonomous vehicles, it becomes apparent that despite their best efforts not all issues could have been contemplated in advance, the result of which is applying laws which were not made with autonomous vehicles in mind. When these vehicles become more prominent in the UK these same issues will present themselves. There are insufficiencies in UK law that ought to be addressed in anticipation of autonomous cars becoming commercially available in an effort to reduce the number of problems that will subsequently arise. Looking to states such as California provide guidance for the difficulties faced when attempting to legislate for these new vehicles. What can be seen most prominently is the underlying complexities within what appears to be a reasonably straightforward divide between the operator’s and manufacturer’s duties, and accordingly, careful consideration will be required to ensure the law is not unduly burdensome on either manufacturers or consumers.










[1] S. Gibbs, ‘Self-Driving Cars: Who’s Building Them and How Do They Work?’, The Guardian, 26 May 2016

[2] T. Adams, ‘Self-Driving Cars: From 2020 You Will Become a Permanent Backseat Driver’, The Guardian, 13 September 2015

[3] The Guardian, ‘Consumer Reports Urges Tesla to Disable Autopilot After Driver’s Death’, 14 July 2016

[4] Revised California Express Terms, Title 13, Division 1, Chapter 1 Article 3.7 – Autonomous Vehicles, revised 30 September 2016, §227.86 (a)(1)

[5] F. Douma and S. A. Palodichuk, ‘Criminal Liability Issues Created by Autonomous Vehicles’ 2012 Santa Clara Law Review vol.52 1157, p.1161

[6] D. Yadron and D. Tynan, ‘Tesla Driver Dies in First Fatal Crash While Using Autopilot Mode’, The Guardian, 1 July 2016

[7] J. Kiss, ‘Tesla has No Plans to Disable Autopilot Mode as Third Recent Crash is Revealed’, The Guardian, 12 July 2016

[8] Revised California Express Terms §227.86 (b)

[9] Draft California Express Terms, Title 13, Division 1, Chapter 1 Article 3.7 – Autonomous Vehicles, revised 16 December 2015, §227.84 (d)

[10] R. A. Cohen; ‘Self-Driving Technology and Autonomous Vehicles: A Whole New World for Potential Product Liability Discussion’ Defense Counsel Journal, July 2015 328, p.329

[11] Douma and Palodichuk; ‘Criminal Liability Issues’ p.1160

[12] A. Herd; ‘R2DFord: Autonomous Vehicles and the Legal Implications of Varying Liability Structures’ Faulkner Law Review vol.5 29, p.40

[13] RTA s.2A, 3ZA

[14] Ibid s.41D(a)

[15] Daborn v Bath Tramways Motor Co Ltd and Smithy [1946] 2 All ER 333

[16] S. Levin and N. Woolf; ‘Tesla Driver Killed While using Autopilot was Watching Harry Potter, Witness Says’, The Guardian, 1 July 2016

[17] Herd, ‘R2DFord’ p.38

[18] Ibid

[19] Ibid p.44

[20] CPA s.3(2)(a)

[21] Ibid s.3(2)(b)

[22] G. E. Marchant and R. A. Lindor; ‘The Coming Collision Between Autonomous Vehicles and the Liability System’ 52 Santa Clara Law Review 2012 1321, p.1323-4

[23] Herd; ‘R2DFord’ p.47-8

[24] Cohen; ‘Self-Driving Technology’ p.333

[25] Donoghue v Stevenson 1932 AC 562

[26] Ibid p.578-9

[27] Evans v Triplex Safety Glass co Ltd [1936] 1 All ER 283

[28] Grant v Australian Knitting Mills [1936] AC 85 p.101

[29] CPA s.2

[30] [2000] PIQR P164

[31] Cf. Bolton v Stone [1951] AC 851 where the risk of injury was so remote it was not sufficient to establish negligence

[32] [2002] EWHC 490 (QB)

[33] Ibid, para.60

[34] As per Daborn. See note 15



































[1] Vettas, N, ‘Developments in Vertical Agreements’, The Antitrust Bulletin (2010) 55(4), 843-874.

[2] Bork, R, The Rule of Reason and the Per Se Concept: Price Fixing and Market Division – Part II’, (1966) ,Yale L.J, 75(3), 377- 475

[3] Jones, A and Sufrin, B (2016). EU Competition Law. 6th ed. Oxford: Oxford University Press


[4] Comanor, W, ‘Vertical Price-Fixing, Vertical Market Restrictions, and the New Antitrust Policy’, (1985), Harvard L.R, 98(5), 983-1002

[5] Paldor, I, ‘The Vertical Restraints Paradox: Justifying the Different Legal Treatment of Price and Non- Price Vertical Restraints’, (2008) The University of Toronto L.J, 58(3), 317-353

[6] Telser, L, ‘Why Should Manufacturers want Fair Trade?, (1960), Journal of Law and Economics 3(1), 86-105

[7] Paldor, I, ‘The Vertical Restraints Paradox: Justifying the Different Legal Treatment of Price and Non- Price Vertical Restraints’

[8] Zevgolis, N, ‘Resale Price Maintenance (RPM) in European competition law: legal certainty versus economic theory?’ (2013), E.C.L.R 34(1), 25-32

[9] Marvel, H.P, ‘Resale Price Maintenance and Resale Prices: Paying to Support Competition in the Market for Heavy Trucks’(2010), Antitrust Bulletin 55(2), 79-99

[10] Leegin Creative Leather Products Inc v PSKS Inc (1997) 522 US 3

[11] Rey, P and Stiglitz, J,  ‘The Role of Exclusive Territories in Producers Competition’, (1995), RAND Journal of Economics ,26(3), 431

[12] Guidelines on vertical restraints [2010] O. J. C 130/1 , para. 106-109

[13] Consumer Focus, Consumer Focus Response to Vertical Restraints Block Exemption Regulation, (September 2009), 11-12

[14] European Commission, Green Paper on Vertical Restraints in EC Competition Policy, COM (96) 721, para. 54

[15] Kneepkens, M,’Resale Price Maintenance: Economic Call for A More Balanced Approach’, (2007) 28(12) E.C.L.R, 660-661

[16] Comanor, W, ‘Vertical Price-Fixing, Vertical Market Restrictions, and the New Antitrust Policy, Harvard Law Review’

[17] Herbert J. Hovenkamp. Economics and Federal Antitrust Law. St. Paul, MN: West Publishing Co. 1985, xvii, 592 pages (Hornbook Series Lawyer’s Edition);  Jay Palmer v BRG of Georgia (2010) 498 U.S. 46 illustrates the reality of a price increase by horizontal territorial market division

[18] B. Durand, “On the Efficiency of VTR” (thesis, Boston College, The Department of Economics, U.S.A., May 2000).

[19] Iacobucci, E, ‘The Case for Prohibiting Resale Price Maintenance’, (1995) 19(2) World Comp.L. & Econ.Rev. 71

[20] Sullivan, E.T and Jeffrey, L (2014). Understanding Antitrust and Its Economic Implications . 6th ed. Newark: Mathew Bender & Co. 227

[21] Jedličková, Barbora (2012) The law of vertical territorial and price restraints in the EU and in the USA: a critical analysis of vertical territorial and price restraints - an argument against legalisation. PhD thesis.

[22] Organisation for Economic Co-operation and Development (‘OECD’), Competition Policy and Vertical Restraints: Franchising Agreements (1994) 192–3. The EC believes that inter-brand competition is a crucial indicator of workable competition: European Commission, Green Paper on Vertical Restraints in EC Competition Policy, COM(96)721, 20

[23] European Commission, Green Paper on Vertical Restraints in EC Competition Policy, COM(96) 721, para. 70 and 78

[24] Comanor, W, ‘Vertical Arrangements and Antitrust Analysis’, (1987) 62(5), N.Y.U L. Rev. 1153

[25] Comanor, W, ‘Vertical Price-Fixing, Vertical Market Restrictions, and the New Antitrust Policy’

[26] Bennett, M et al., Resale Price Maintenance: Explaining the Controversy, and Small Steps Towards a More Nuanced Policy, (2010), MPRA Paper No. 21121, Online at [Accessed 20 December 2016]

[27] Treaty on the Functioning of the European Union, O.J. C 83 of 30.3.2010 

[28] Regulation 2790/99 Vertical Restraints [1999] O.J. L336/21

[29] Guidelines on Vertical Restraints [2000] OJ C291/1

[30] Jones, A and Sufrin, B (2016). EU Competition Law, 768

[31] Guidelines, para 110

[32] Commission Regulation 330/2010 on the application of Article 101(3) of the Treaty

on the Functioning of the European Union to categories of vertical agreements and

concerted practices, O.J. L 1021

[33] Brenning –Louko, M et al., ‘Vertical Agreements: New Competition Rules for the Next Decade’, (2010) 2 Competition Policy Newsletter, 15

[34] Notice on agreements of minor importance which do not appreciably restrict

competition under Article 81(1) of the Treaty establishing the European Community (“de minimis Notice”) [2001] OJ C 368/13

[35] Case 26/76 Metro – SB – Großmärkte GmbH v Commission (Metro I), [1977] ECR 1875, para.21

[36] Jones, A,. ‘Resale Price Maintenance: A Debate About Competition Policy in Europe’, (2009), European Competition Journal, 5(2), 479-514

[37] Case C -167/04 JCB Service v Commission [2006] ECR I-8935; Yamaha, IP/03/1028, 16 July 2003

[38] Bennett, M et al., Resale Price Maintenance: Explaining the Controversy, and Small Steps Towards a More Nuanced Policy, (2010)

[39] Van Doorn, F, Resale Price Maintenance in EC Competition Law: The Need for  Standardised approach, (November 6 2009). Online at [Last Accessed 23 December 2016]

[40] Case C-74/04 P Commission v. Volkswagen AG [2006] ECR I-6585

[41] Cases 56/64, 58/64 Établissements Consten S.à.R.L. and Grundig-Verkaufs-GmbH v. Commission of the European Economic Community [1966] ECR 299

[42] Case 56/65, Société La Technique Minière v Maschinenbau Ulm GmbH [1966] ECR 235,

[43] Yamaha, IP/03/1028, 16 July 2003

[44] Case C -167/04 JCB Service v Commission [2006] ECR I-8935

[45] Guidelines, para 61

[46] IP/02/916, Commission clears B&W Loudspeakers distribution system after company deletes hard-core violations, 24 June 2002

[47] COMP/3344 Grundig, 23 September 1964

[48] Case T-13/03, Nintendo and Nintendo Europe v Commission [2009] ECR II-975

[49] Joined cases C-501/06 P, C-513/06 P and C-519/06 P GlaxoSmithKline Services Unlimited v Commission of the European Communities [2009] I-09291

[50] A similar conclusion was drawn in Case 243/83 Binon & Cie v SA Agence et Messageries de la Presse [1985] ECR 2015

[51] Jones, A and Sufrin, B (2016). EU Competition Law. 817

[52] Guidelines, para 47

[53] Maci, M, ’The assessment of RPM under EU Competition rules: certain inconsistencies based on a non-substantive analysis’, (2014) E.C.L.R, 35(3), 103-109

[54] Kyprianides, G.P, ‘Should Resale Price Maintenance be per se illegal?’, (2012), E.C.L.R, 33(8),  376-385

Go Back