A Look at the Legal Environment for Driverless Vehicles (2016)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

41 V. CRIMINAL LAW AND PROCEDURE Driverless vehicles will develop a complicated, mul- tifaceted relationship with federal and state criminal law and procedure.336 These devices may lead to the recognition of new crimes, even as they reduce the overall number of traffic infractions and other crimes committed with automobiles. They also may enhance the surveillance capabilities of the government, even as they diminish the number of traffic stops that, today, represent the most common form of interaction between police officers and the general public.337 These developments will take time to unfold. More immediately, driverless vehicles will inherit a large body of existing laws that address the proper operation of motor vehicles, their required equip- ment, and the misuse or exploitation of these devices. Many of these laws attach criminal penalties to vio- lations of their provisions. Some of these existing statutes admit to ready application to driverless vehicles. Other laws may require revisions to clarify how they will apply, if at all, to this new form of transportation. And still other laws, presently not in existence, may impose new criminal prohibitions that specifically address novel antisocial behaviors associated with driver- less vehicles. The future also may see a reevaluation of who, or what, should be held criminally liable for crimes involving driverless vehicles. As discussed in con- nection with this report’s discussion of civil liability, the automation of functions associated with the operation of driverless vehicles will entail a shift in responsibility from human drivers to the automo- biles themselves and the systems they rely upon. Just as this shift may lead to the reconsideration of who may be held liable in a civil action for damages after an accident involving a driverless vehicle, it also may call into question existing assignments of criminal responsibility. A. Criminal Laws Federal and state laws relate hundreds of crimes pertinent to motor vehicles. A basic taxonomy would some observers presently predict,334 and unless a transformative change in law occurs, at least some lawsuits against manufacturers will persist. Yet the anticipated overall decline in personal injury litigation associated with automobiles may have important consequences for a substantial segment of the bar for whom these matters represent a sig- nificant share of their case portfolios.335 These prospects all lie far in the future. Pres- ently, the main issue before policymakers concerns whether to avoid this anticipated gradual change through the near-term enactment of statutes or promulgation of regulations that preempt or other- wise limit tort lawsuits associated with driverless vehicles. If conventional vehicles provide any guide, some preemption of tort liability vis-à-vis basic safety features and certain programming choices is probably inevitable in the long run. But the infor- mation required for the adoption of sound, long- term regulatory standards has not yet been gener- ated, and broad preemption has not yet been necessary for innovation to occur in this field. Fur- thermore, the incremental and ongoing develop- ment of automated and driverless vehicle technolo- gies militates against premature regulatory strategies—of any stripe—that may be overbroad or off the mark, and prove difficult to amend at a later date. Finally, policymakers should appreciate that the civil liability law that comes to surround driverless vehicles will itself serve as a foundation for princi- ples later extended to other applications of artificial intelligence, if and when these technologies begin to cause harm. To date, software has not left a particu- larly large footprint on tort law, primarily because software programming decisions rarely have led to personal injuries. That will change once driverless vehicles become common. It will change even more if and when robots and other products involving sophisticated artificial intelligence become more prevalent and useful. The tort law that coalesces around driverless vehicles will represent the open- ing chapter of a longer narrative in which our soci- ety decides where and how to address and account for risks associated with these devices. 334 But cf. RAND rePorT, supra note 181, at 118 (“Overall, we do not anticipate that liability for individual drivers will be a problematic obstacle or deterrent to the use of AV tech- nologies.”); id. at 132 (“The existing liability regime does not seem to present unusual liability concerns for owners or drivers of vehicles equipped with AV technologies.”). 335 See Erik Turkewitz, Will Google Cars Eviscerate the Personal Injury Bar?, N.Y. PersonAl inJUry lAW blog (Dec. 22, 2014), http://www.newyorkpersonalinjury attorneyblog.com/2014/12/will-google-cars-eviscerate-the- personal-injury-bar.html. 336 Other authors also have considered the topic addressed in this section. E.g., Frank Douma & Sarah Aue Palodichuk, “But Officer, it Wasn’t My Fault…the car did it,” Criminal Liability Issues Created by Autonomous Vehicles, 52 sAnTA ClArA l. rev. 1157 (2012); Jeffrey K. Gurney, Driving into the Unknown: Examining the Cross- roads of Criminal Law and Autonomous Vehicles, 5 WAke foresT J. lAW & Pol’y 393 (2015). 337 ChrisTine eiTh & mATTheW dUrose, ConTACTs beTWeen PoliCe And The PUbliC, 2008, U.S. Department of Justice, Office of Justice Programs, Special Report NCJ234599, 1 (Oct. 2011).

42 1. Reassignment and Recasting of Criminal Liability In some contexts, driverless vehicles eventually may lead to the reassignment of criminal liability from its current bearer to someone (or something) else, or to the replacement of low-level sanctions with other methods of deterrence and punishment. For example, current state laws typically make the “driver” or “operator” of a vehicle liable for failing to stop at a stop sign.339 If a driverless vehicle fails to obey such a sign, conceivably the human user of the vehicle could continue to be held liable. To reach this result, however, the user either would have to be considered the “driver” or “operator” of the vehicle, or the pertinent statute would have to be amended to make mere users responsible for their actions. Continued assignment of liability to users seems likely insofar as Level 3 vehicles are involved.340 That said, the prospect of criminal lia- bility may undermine key benefits associated with driverless vehicles, especially as these devices grow safer and the need for human oversight diminishes. As time passes and the public grows more comfortable with driverless vehicles, govern- ments may find it appropriate to revisit the impo- sition of criminal liability on passive users of advanced driverless vehicles.341 In its place, gov- ernments could fine or impose other penalties upon the manufacturers of vehicles that commit traffic violations, or upon the makers of software or other components that direct these devices; they could simply remove these vehicles from operation; or they could adopt some other response. Debates over whether to shift responsibility for traffic infractions from users to third parties may be contentious, however. Prevailing notions regarding the responsibility of drivers and users for their auto- mobiles may prove difficult to dislodge. It also may be unclear for some time who, as among users and various product manufacturers, bears responsibility for various aspects of a vehicle’s performance. This confusion may counsel against the reassignment of liability, at least for some time. It is possible that an interregnum period will appear in which drivers will be permitted to claim a defense by ascribing an improper traffic maneuver to a self-driving function, but no direct liability will attach to the manufac- turer or anyone else for this error. distinguish among basic regulatory offenses (e.g., equipment violations and simple deviations from rules of the road); crimes that address aggravated forms of user misconduct (e.g., driving under the influence of alcohol, hit-and-run); specific forms of more general crimes (e.g., automobile theft, carjack- ing); and criminal conduct that are facilitated by or otherwise connected to motor vehicles, but as to which the pertinent criminal prohibition does not specifically reference or require these devices (e.g., transportation of narcotics for purposes of sale). Crimes within the first of these categories tend to carry modest penalties and minimal mens rea (intent) requirements. The other types of crimes run the gamut from infractions punishable by only a fine to serious felonies that can carry long prison terms. Some, though not necessarily all, of these offenses will apply to driverless vehicles, with little need for translation or modification. For example, many of the basic equipment requirements found within state vehicle codes, such as the common requirement that a vehicle have operating headlamps, presumably will apply both to driverless vehicles and to vehicles with active human drivers. There may come a time when the driverless vehicles are so predominant on the roads and their sensory capabilities so robust that a subset of these safety laws may become obsolete. But until then, it is expected that these rules will con- tinue in force, and will be of general application. Other provisions of state vehicle codes may require some modifications to account for the distinct charac- teristics of driverless vehicles, as related below.338 338 As a threshold matter, state vehicle codes tend to pre- sume that vehicles have a “driver.” To this point, the exis- tence and nature of a “driver” both have been sufficiently obvious that states have not provided a robust definition of this term. With driverless vehicles, this will change. One leading commentator has observed that at present, the ambiguity inherent in the word “driver,” as that term is used in various contexts, means that driverless vehicles are “prob- ably” legal in the United States, but under present law “there is likely to be some person connected to the vehicle who must be licensed, who may need to be physically present, and who must act prudently.” Smith, Automated Vehicles Are Prob- ably Legal in the United States, supra note 180, at 480 (foot- notes omitted). While some individuals and entities may be comfortable with these ambiguities, others may not be, and will continue to press for laws that clarify how a driverless vehicle may be utilized without fear of criminal sanctions. In this vein, some states already have enacted laws that permit road testing of autonomous vehicles, under certain conditions. On the issue of compliance with traffic laws, regu- lations promulgated in Nevada prescribe that “a person shall be deemed the operator of an autonomous vehicle which is operated in autonomous mode when the person causes the autonomous vehicle to engage, regardless of whether the per- son is physically present in the vehicle while it is engaged,” nev. Admin. Code § 482A-020(2) (2014), and “For the purpose of enforcing the traffic laws and other laws applicable to drivers and motor vehicles operated in this State, the opera- tor of an autonomous vehicle that is operated in autonomous mode shall be deemed the driver of the autonomous vehicle regardless of whether the person is physically present in the autonomous vehicle while it is engaged.” Id., § 482A-030(2). 339 E.g., CAl. veh. Code § 22450 (West 2000). 340 See Gurney, supra note 336, at 415–16. 341 Id.

43 state might work toward this goal by permitting the use of a Level 3 vehicle in situations where it would be impermissible to drive or operate a vehicle that requires more human engagement. But other states may regard the ability of human user to reassert con- trol over even a Level 3 vehicle as a sufficient basis for a DUI offense. If states vary in their responses, experiences under different regimes likely will inform the approaches taken to this issue when more advanced, Level 4 vehicles appear. 4. Generic Crimes Driverless vehicles also may make it easier to commit other crimes not inherently connected to automobiles.344 Their potential for use as unmanned “drones” suggests that they could be deployed for terrorist purposes. Terrorists or other third parties also may try to “hack” individual vehicles or the sys- tem in which these devices operate—either to trig- ger collisions or simple confusion, or to gain access to data compiled by driverless cars for purposes of surveillance or profit. It is unclear whether new criminal laws will be necessary to address these hypothetical scenarios. Most third-party criminal behavior involving driv- erless cars already is captured by existing crimes, such as laws that prohibit computer “hacking”345 or the unauthorized transportation of explosive devices.346 Nevertheless, the existence of generic crimes will not necessarily sate public demand for an additional criminal prohibition to serve as a direct, firm response to a novel threat. The rela- tively recent adoption of “carjacking” laws offers a case in point. Although the conduct associated with carjacking amounts to robbery, a crime of ancient vintage, in the early 1990s several states and the federal government responded to amplified fears about this sort of offense by enacting special car- jacking statutes that attached heightened penal- ties to the use of force to commandeer an automo- bile.347 Alternatively, use of driverless vehicles in the commission of a crime may be made the subject of a criminal “enhancement,” parasitic to an exist- ing offense, whereby a stiffened penalty may be imposed when a crime is committed using a driver- less vehicle. 2. Rules of the Road As the stop sign example above indicates, although driverless vehicles will be subject to most or all of the prevailing “rules of the road,” they also may occasion changes or additions to existing directives. Some of these shifts likely will precede the broad diffusion of driverless vehicles, while others will evolve gradually, over the course of many years. As previously discussed, with automobiles the first rules of the road and equip- ment requirements related to registration and licens- ing, permissible spheres of usage, etiquette when interacting with other forms of transportation, the necessary (basic) safety equipment, and speed limits. Other, more nuanced rules took longer to appear. A similar response to driverless vehicles (already incipi- ent in those states that have permitted the limited road testing of these devices) would include, early on, such basic matters as vehicle registration, licensing (including driver qualifications), areas in which driver- less vehicles may be used, necessary conduct when interacting with conventional vehicles, insurance requirements, and a clarification of who serves as a “driver,” “operator,” or controller of these devices. Other rules would appear later, when proper policy would hinge on the particular path that this technology has taken in its development and adoption. 3. Other Driving Offenses Other existing crimes likewise may require modifi- cations to account for how driverless vehicles are used. Hit-and-run statutes, for instance, commonly require the “driver” of a vehicle to stop and give both informa- tion and assistance at the site of an accident.342 A driv- erless vehicle may be fully capable of transmitting the information required by these laws. But direct human engagement may be required for the provision of med- ical assistance. These statutes therefore may have to be rewritten or supplemented to assign greater responsibilities to “users” or “occupants” of driverless vehicles in the event of an accident. Other traffic laws that impose similar duties on drivers or operators may require comparable adjustments. These changes may occur along different timeta- bles across the states. With the crime of driving under the influence of alcohol, for example, the appearance of driverless vehicles may lead some states to amend their DUI laws (which commonly prohibit “driving” or “operating” a motor vehicle while in a prohibited state or degree of intoxication) to encourage poten- tially inebriated drivers to avail themselves of a driv- erless option, such as a for-hire driverless taxi.343 A 342 CAl. veh. Code §§ 20001, 20003(a) (West 2000 & Supp. 2015). 343 See Gurney, supra note 336, at 421–23 (discussing this possibility). 344 Douma & Palodichuk, supra note 336, at 1159. 345 Every state has some form of computer-crime law on the books. Jordan M. Blanke, Criminal Invasion of Pri- vacy: A Survey of Computer Crimes, 41 JUrimeTriCs 443, 449 (2001). 346 E.g., 18 U.S.C. § 842 (2012). 347 E.g., Anti Car Theft Act of 1992, Pub. L. No. 102-519, 106 Stat. 3384; Act of Sept. 30, 1993, ch. 611, 1993 Cal. Stat. 3456.

44 6. Federal Crimes The discussion above presumes that states will continue to occupy the lead role in crafting and enforcing criminal laws applicable to motor vehicles operated on the highways. That said, the federal government also has played a role in the creation and enforcement of crimes involving motor vehicles. Back in 1919, Congress passed the Dyer Act, also known as the National Motor Vehicle Theft Act, which made interstate motor vehicle theft a federal crime.351 This law provided federal prosecutors with a steady stream of cases for several decades. More recently enacted federal crimes applicable to auto- mobiles, such as the odometer fraud provisions of the Motor Vehicle Information and Cost Savings Act of 1972352 and the carjacking crime within the Fed- eral Anti Car Theft Act of 1992,353 today also produce a small number of federal prosecutions. More com- monly, federal prosecutors pursue generic crimes that can be facilitated by automobiles, such as the interstate transportation of narcotics. Driverless vehicles may lead to new federal crimes. The Commerce Clause of the United States Constitution confers upon Congress the power to regulate interstate commerce. Courts have held that the federal carjacking statute represents a valid exercise of this authority,354 paving the way for fed- eral crimes that would address interference with an increasingly networked highway system. Yet when and how the federal government will exercise this authority is at this point unclear. In this legal con- text, as in other spheres, the circumstances sur- rounding the development and spread of driverless vehicles likely will influence the timing and nature of federal engagement with these devices. B. Criminal Investigations If they become common, driverless vehicles also will have an important effect on police investigations. These devices will alter the existing “equilibrium” between privacy and security by enabling both new crimes and new methods of surveillance.355 Among 5. New Crimes In addition to enhancing existing criminal capa- bilities, driverless vehicles also may inspire rela- tively unprecedented forms of misconduct, unique to the technology, that in turn generate new criminal laws. The recent spate of state laws that criminalize “sexting” and “revenge pornography” illustrate that although some time may pass before it becomes apparent that a new technology is being utilized in a manner that compels the creation of a new crime,348 legislatures can respond quickly to such concerns once they arise.349 One type of new crime that may become associated with driverless vehicles would involve user modification of the software installed in these devices. It has been observed that just as the users of cellular phones may “hack” their devices to enhance their capabilities, users of driverless vehi- cles likewise may try to override functions within their vehicles that limit the speeds at which they operate, or control other aspects of their perfor- mance.350 This sort of owner modification of driver- less vehicles may become the subject of criminal liability, as hacking by a third party (or odometer tampering by a vehicle seller) already are. Yet these concerns about new criminal capabilities remain speculative at this point. One lesson that emerges from innovations of the past is that criminal laws enacted at an early phase of a technology’s development often prove stubbornly resistant to change, even as the technology grows more sophisti- cated, patterns of use and misuse shift, vocabularies surrounding the technology change, and a different policy response becomes increasingly desirable. This intransigence can be particularly acute when crimes have been enacted and any modification would reduce the scope of an offense or lessen the attached punish- ment. The “stickiness” of early adopted policies tends to counsel in favor of significant deliberation prior to the recognition of any stringent and severe new crim- inal laws concerning a new technology. 348 The first law specifically prohibiting “obscene phone calls,” for example, appears to have been enacted in 1907, decades after telephones first appeared. An Act Prohibiting the Uttering of Lascivious or Obscene Language over Tele- phones in this State, ch. 249, 1907 N.D. Laws 387. 349 See Sameer Hinduja & Justin W. Patchin, State Sexting Laws, CyberbUllying res. CTr. (2015), http:// cyberbullying.us/state-sexting-laws.pdf (indicating that as of January 2015, 20 states had enacted laws that pro- hibit “sexting”). As of April 2015, 18 states have enacted “revenge porn” laws that criminalize the nonconsensual online posting of sexually explicit images. Alex Ronan, Could All of These New Revenge-Porn Laws Actually Be a Bad Thing?, N.Y mAg. (April 16, 2015), http://nymag. com/thecut/2015/04/why-regulating-revenge-porn-is-so- tricky.html. 350 RAND rePorT, supra note 181, at 70–71. 351 National Motor Vehicle Theft Act, Pub. L. No. 66-70, 41 Stat. 324 (1919). 352 Motor Vehicle Information and Cost Savings Act, Pub. L. No. 92-513, §§ 403–06, 86 Stat. 947, 961 (1972). 353 Anti Car Theft Act of 1992, Pub. L. No. 102-519, 106 Stat. 3384. See also F. Georgann Wing, Putting the Brakes on Carjacking or Accelerating It – The Anti-Carjacking Act of 1992, 28 U. riCh. l. rev. 385, 410–22 (1994) (discussing the legislative history of this law). 354 E.g., United States v. Runyon, 707 F.3d 475, 489-90 (4th Cir. 2013). 355 See Orin S. Kerr, An Equilibrium-Adjustment Theory of the Fourth Amendment, 125 hArv. l. rev. 478 (2011) (describing how new technologies may disrupt this equilibrium).

45 while another five (one justice joining both camps) would have held that the prolonged use of the GPS device under the specific circumstances presented (involving an investigation of suspected narcotics trafficking) violated the defendant’s reasonable pri- vacy expectations.362 Although Jones signifies that a majority of the present Court apparently regards trespassory and some non-trespassory surveillance of vehicle move- ments as implicating the Fourth Amendment, it remains unclear precisely when the government’s non-trespassory acquisition of information regard- ing a driverless vehicle’s whereabouts will be treated as a “search.” With driverless vehicles, this collection could occur along different avenues. The government may try to obtain this information through infrastructure built for connected-vehicle, including V2I, purposes, or through a more basic repurposing of existing surveillance tools such as cameras and license plate readers positioned on and about the highways.363 Law enforcement also may invoke the “third-party doctrine,” under which an individual lacks a reasonable expectation of pri- vacy in information that he or she voluntarily com- municates to a third party,364 to justify the warrant- less acquisition of information regarding driverless-vehicle use from communications pro- viders or from companies that manufacture or maintain driverless vehicles. Presently, one cannot predict with certainty how these warrantless-surveillance scenarios will be resolved. At least one member of the U.S. Supreme Court has expressed reservations about the contin- ued application of the third-party doctrine,365 and it is possible that this rule and related Fourth Amend- ment doctrine will change before driverless vehicles become common. Searches of driverless vehicles themselves, as opposed to mere surveillance of their movements, also may raise interesting legal issues. Automobiles represent “effects” under the Fourth Amendment to the United States Constitution, and their owners and possessors generally can claim a reasonable expectation of privacy as against physical intrusions their privacy consequences, driverless vehicles will collect a tremendous amount of information regard- ing their users’ movements, information that law enforcement may want to obtain without a search warrant. It is anticipated that these efforts will be challenged by defendants and others who regard such efforts as impermissible under the Fourth Amendment to the United States Constitution.356 These arguments will implicate what is presently a hotly contested area of law. Under the Fourth Amendment, the government engages in a “search” either when it intrudes upon an individual’s subjec- tive expectation of privacy, to the extent that society is prepared to regard such an expectation as reason- able,357 or when it commits an information-gather- ing “trespass” upon a person or their papers, houses, or effects.358 For a “search” to be lawful, the govern- ment first must have obtained a search warrant that authorizes the search, or a recognized exception to the warrant requirement must apply. A search warrant can be obtained only on a showing of prob- able cause, a standard equated with a “fair probabil- ity” that the search will yield evidence of a crime.359 A recent decision by the United States Supreme Court regarding prolonged government surveillance of vehicle movements yielded a consensus that GPS surveillance (at least on the facts presented) amounts to a “search” that triggers Fourth Amend- ment protections, but a split over the analytical approach that compels that conclusion. In 2012, the Court held, unanimously, in United States v. Jones that a “search” occurred when government agents placed a GPS device on a suspect’s automobile and then used that device to track the suspect’s move- ments for 28 days.360 Significantly, however, the jus- tices did not agree on a single rationale for this hold- ing. Five justices relied upon the “trespass” theory,361 356 The Fourth Amendment provides: The right of the people to be secure in their per- sons, houses, papers, and effects, against unreason- able searches and seizures, shall not be violated, and no warrants shall issue, but upon probable cause, sup- ported by oath or affirmation, and particularly describ- ing the place to be searched, and the persons or things to be seized. U.S. ConsT. amend. IV. 357 Katz v. United States, 389 U.S. 347, 361, 88 S. Ct. 507, 19 L. Ed. 2d 576, 587 (1967) (Harlan, J. concurring). 358 See, e.g., United States v. Jones, __ U.S. __, 132 S. Ct. 945, 949, 181 L. Ed. 2d 911 (2012) (holding that the government engages in a “search” when it physically occupies a person’s vehicle for the purpose of obtaining information). 359 Illinois v. Gates, 462 U.S. 213, 238, 103 S. Ct. 2317, 2332, 76 L. Ed. 2d 527, 548 (1983). 360 Jones, 132 S. Ct. at 949. 361 Id. 362 Id. at 957–64 (Alito, J., concurring). 363 Devlin Barrett, U.S. Spies on Millions of Drivers, The WAll sTreeT JoUrnAl, Jan. 26, 2015, at A1. 364 See Smith v. Maryland, 442 U.S. 735, 743-44, 99 S. Ct. 2577, 2581-2584 (1979); United States v. Miller, 425 U.S. 435, 443, 96 S. Ct. 1619, 1624, 48 L. Ed. 2d 71, 79 (1976). See also Orin S. Kerr, In Defense of the Third- Party Doctrine, 107 miCh. l. rev. 561, 563 (2009) (discussing this doctrine and critiques thereof). 365 See Jones, 132 S. Ct. at 957 (Sotomayor, J., concur- ring).

46 There also exist other discrete circumstances in which vehicles may become subject to government inspection without a search warrant, such as when a vehicle enters the country through an interna- tional border.374 In many circumstances, these warrant exceptions will apply to driverless vehicles just as they do to conventional automobiles. In other cases, however, the nature of the evidence that law enforcement offi- cers may seek to obtain from driverless vehicles may complicate the necessary Fourth Amendment analy- sis. In this respect, the culling of digital evidence from driverless vehicles by police may prove particu- larly sensitive. Recently, in Riley v. California,375 the U.S. Supreme Court barred the warrantless search of cellular telephones incident to the arrests of their possessors.376 In reaching this result, the Court emphasized the quantity and potentially sensitive nature of the digital information in these devices.377 As applied to driverless vehicles, this ruling will encourage defendants to challenge the government’s warrantless acquisition of digital information from their vehicles, even when this action is taken pursu- ant to a recognized warrant exception. A defendant in such a case could argue with some force that the collection of broad swaths of location data or other information concerning vehicle use pursuant to, for example, the automobile exception is just as unrea- sonable as the warrantless search of a cell phone incident to arrest was deemed to be in Riley.378 Finally, and perhaps most intriguingly, the emer- gence of driverless vehicles may function to limit police interactions with the public. If driverless vehicles lead to fewer traffic violations and impaired driving offenses, there likely will be fewer occasions for police to pull motorists over. This diminution of traffic stops could have profound consequences for police staffing and deployment, and on the use of traffic stops to enforce other substantive criminal prohibitions. C. Conclusions The criminal laws that will become attached to driverless vehicles will not be written on a blank slate. These devices will be subject to a large body of existing law applicable to conventional motor vehi- cles. And between now and when driverless vehicles by the government.366 Therefore, a police officer’s entrance into and search of a vehicle amounts to a “search” that requires a warrant or warrant exception. Several such exceptions apply to automobiles. These exceptions are premised on rationales includ- ing the pervasive regulation of these devices, their susceptibility to movement, and the fact that they tend to be operated in public places.367 With regard to automobiles, the most commonly invoked warrant exceptions consist of: • Consent: Police may search a vehicle if they receive voluntary consent to search from someone with actual or “apparent” authority.368 • The “automobile exception”: Police also may search a vehicle, as thoroughly as they could if they had a search warrant, if there is probable cause that the vehicle contains contraband or evidence of a crime.369 • Searches incident to arrest: Police may search the area of a vehicle within grabbing distance of a lawfully arrested recent occupant, and may search the interior of a vehicle’s passenger compartment if it is reasonable to believe that the vehicle contains evidence of the offense of arrest.370 • Vehicle frisks: Police may search the passenger compartment of a vehicle for a weapon if the officer has reasonable suspicion that the vehicle contains a weapon, and that an occupant of the vehicle may become armed and dangerous if allowed access to the weapon.371 • Inventory searches: A lawfully impounded vehi- cle may be physically searched by law enforcement pursuant to preexisting inventory policies;372 and • Regulatory searches: Under limited circumstanc- es, police may search a vehicle to ensure compliance with basic registration and regulatory requirements.373 374 United States v. Flores-Montano, 541 U.S. 149, 155, 124 S. Ct. 1582, 158 L. Ed. 2d 311, 318 (2004). 375 __ U.S. __, 134 S. Ct. 2473, 189 L. Ed. 2d 430 (2014). 376 Id. at 2485. 377 Id. at 2489–91. 378 See Dorothy J. Glancy, Sharing the Road: Smart Transportation Infrastructure, 41 fordhAm Urb. L.J. 1617, 1649–51 (2014) (discussing the Riley case and its implications for driverless vehicles). 366 See California v. Carney, 471 U.S. 386, 391–92, 105 S. Ct. 2066, 2069-2070 (1985) (discussing the premises behind the “automobile exception” to the warrant requirement); Cardwell v. Lewis, 417 U.S. 583, 590, 94 S. Ct. 2464, 2469, 41 L. Ed. 2d 334 (1974) (same). 367 Id. 368 Florida v. Jimeno, 500 U.S. 248, 251, 111 S. Ct. 1801, 1803 (1991); Illinois v. Rodriguez, 497 U.S. 177, 188, 110 S. Ct. 2793, 2801 (1990). 369 California v. Acevedo, 500 U.S. 565, 580, 111 S. Ct. 1982, 114 L. Ed. 2d 619, 634 (1991). 370 Arizona v. Gant, 556 U.S. 332, 351, 129 S. Ct. 1710, 1723–1724, 173 L. Ed. 2d 485, 501 (2009). 371 Michigan v. Long, 463 U.S. 1032, 1049, 103 S. Ct. 3469, 3480-3481, 77 L. Ed. 2d 1201, 1219–1220 (1983). 372 South Dakota v. Opperman, 428 U.S. 364, 372, 96 S. Ct. 3092, 3098–3099, 49 L. Ed. 2d 1000, 1007 (1976). 373 New York v. Class, 475 U.S. 106, 118, 106 S. Ct. 960, 968, 89 L. Ed. 2d 81, 93 (1986).