State Farm Mutual Automobile Insurance Co. v. Kastner 77 P.3d 1256

Justice KOURLIS

delivered the Opinion of the Court.

In this case, an assailant kidnapped Plaintiff, Christina E. Kastner, took her in her own car to a remote location and sexually assaulted her in the vehicle. Kastner sought coverage from her automobile insurer, State Farm Mutual Automobile Insurance Company (State Farm), for her injuries State Farm denied coverage and brought a declaratory judgment action against Kastner, seeking a declaration that its automobile insurance policy did not cover the injuries Kastner had suffered from the assault. The trial court determined that there was coverage, and the court of appeals agreed. State Farm Mut. Auto. Ins. v. Kastner, 56 P.3d 1144 (Colo.App.2002). We granted certiorari on the question of whether injuries caused by a sexual assault in an automobile arise out of the operation, maintenance, or use of a motor vehicle for purposes of personal injury protection or uninsured/underinsured automobile insurance coverage.

Resolution of this case requires us to determine whether the injuries associated with the sexual assault are causally related to a "use" of the claimant's motor vehicle We now hold that (1) where the motor vehicle is being used in a manner reasonably foreseeable at the time the parties contracted for the insurance and (2) the "use" of the vehicle is inextricably linked to the plaintiff's injury, the plaintiff is entitled to recover. Both because we conclude that the use was not reasonably foreseeable and because we con-elude that the sexual assault had an insufficient causal nexus with use of the vehicle, we now hold that Kastner's State Farm policy did not cover her injuries. Accordingly, we reverse the court of appeals, and return this case to the trial court with directions to enter summary judgment on State Farm's motion.

I. Facts

This case was submitted on stipulated facts. Those facts disclose that on December 8, 1998, Christina Kastner was shopping at the Citadel Mall in Colorado Springs. When she left the store sometime after 7:30 p.m., it was dark. Her car was parked about ten cars down from the closest parking spot on the east side of the shopping center. The lot was relatively full and there were cars parked on both sides of Kastner's car. Kast-ner was standing at her car and had unlocked the car door when she saw a male just to the south of her and directly behind her car. She then opened the door to the car and was standing between the car seat and the open door when the male began asking for directions. Kastner believed that the male had been hiding either behind and to the side of her vehicle or behind the vehicle next to hers because she did not notice him until she reached her car.

Kastner had no reason to believe that the male used her vehicle in some way to identify her as a potential victim. As Kastner was responding to the request for directions, the man quickly moved beside her and ordered *1259her to get into the car. She saw something in his hand that she believed to be a knife or gun. Kastner tried to get away from the man by pushing him in the face, but he moved the object toward her and she noticed that the object was a knife. He ordered her to the passenger side of the car, and he grabbed the keys to the car from her hand. She obeyed the man's instructions to get into the front passenger seat of the car, put the seat back and get down as far as possible. The man then entered the driver's side of the car and drove her car from the lot.

He took Kastner to Palmer Park, a wooded park in Colorado Springs. Along the way, he pointed a knife with a 6" to 8" blade at her, yelled at her to stay in the car and threatened to kill her. At Palmer Park, he pulled off the road in an isolated area. He robbed her of $150, and demanded that she disrobe. She opened the passenger door to attempt escape, but was held in by her automatic seatbelts. The man immediately grabbed her by the hair and placed the knife blade on her throat. He then sexually assaulted her in the car with the knife at her back. After the assault, he drove out of Palmer Park and pulled into a liquor store parking lot. He threatened her and her children with bodily harm if she reported the assault, got out of the car and walked away. Kastner immediately drove to the police station to report the incident.

At the time of the incident, Kastner was insured by State Farm Insurance Company under a Personal Injury Protection (PIP) policy that was consistent with the Colorado Auto Accident Reparations Act (the "No Fault Act"), section 10-4-701 et. seq., 8 C.R.S. (2002). The policy also included uninsured/underinsured (UM) motorist coverage pursuant to the Colorado Uninsured Motorist Act, section 10-4-609, 3 C.R.S. (2002). Kast-ner submitted claims to State Farm for both PIP and UM benefits to compensate her for injuries arising out of the subject incident. State Farm denied the claims.

In pertinent part, the personal injury protection portion of that policy provided that:

We will pay in accordance with the No Fault Act for bodily injury to an insured, caused by an accident resulting from the use or operation of a motor vehicle. (emphasis added)

The uninsured motor vehicle portion of that policy provided that:

We will pay damages for bodily injury an insured is legally entitled to collect from the owner or driver of an uninsured motor vehicle. The bodily injury must be caused by an accident arising out of the operation, maintenance or use of an uninsured motor vehicle. (emphasis added)

State Farm brought this action, seeking a declaratory judgment that it had no obligation under the policy. The parties submitted the matter to the trial court on stipulated facts and cross motions for summary judgment. The trial court granted summary judgment for Kastner, concluding that "the victim selection of Kastner after the vehicle door was opened, the use of the reclining passenger seat to prevent Kastner from signaling for help, the use of the vehicle to get to an isolated area and the use of the automatic seatbelts as restraints collectively constitute a causal connection between this vehicle and the assault." State Farm appealed to the court of appeals, which affirmed the trial court, holding that under the facts of this case, there was a sufficient causal connection between the car and the injuries to warrant a finding of coverage under the PIP and UM provisions of the policy. Kastner, 56 P.3d at 1146.

II. Insurance Policies in General

An insurance policy is a contract between the insured and the insurer, and as such, it is to be interpreted according to settled principles of contract law. Allstate Ins. Co. v. Huizar, 52 P.3d 816, 819 (Colo.2002). This general rule, however, has two provisos. First, the contract raises quasi-fiduciary obligations owed by the insurer to the insured. Farmers Group, Inc. v. Trimble, 691 P.2d 1138, 1141 (Colo.1984). Unlike the ordinary commercial contract where the parties seek to ensure a commercial advantage for themselves, the insurance contract "seeks to obtain some measure of financial security and protection against calamity" for the insured. Id. As a result, the insurer has *1260a common-law duty "not to unreasonably withhold payment of benefits it is obligated to make under the insurance contract." Farmers Group, Inc. v. Williams, 805 P.2d 419, 423 (Colo.1991).

Second, the contract must comply with applicable statutory requirements. Should the contract fail to conform to any statute, it is unenforceable to that extent. Peterman v. State Farm Mut. Ins. Co., 961 P.2d 487, 492 (Colo.1998). While the No Fault Act requires the insurer to provide certain coverage within its policy, this requirement co-exists with its quasi-fiduciary duty not to deny this coverage unreasonably. Williams, 805 P.2d at 4283.

Here, we deal with the simple language of a contract, and with the legal implications of that language. Since both the UM and PIP policy provisions are express attempts to conform to statutory requirements, our interpretation of their terms should reflect the overall legislative purpose of the UM and No Fault statutes. See Allstate Ins. Co. v. Parfrey, 880 P.2d 905, 911 (Colo.1992); 7 Lee R. Russ & Thomas F. Segalla, Couch on Insurance § 109:17 (8d ed.1995, updated 2008). Consistent with both the legislative declaration and the general purpose of automobile insurance, our previous cases have interpreted the No Fault Act as providing a floor of recovery for the injuries of policy-carriers for the type of risks one would expect an insurance contract to cover. Seq, eg., Aetna Cas. & Sur. Co. v. McMichael, 906 P.2d 92, 103 (Colo.1995). Thus, the legislative intent of the statutes controls our interpretation of the policies. See State Farm Mut. Auto. Ins. Co. v. McMillan, 925 P.2d 785, 792 (Colo.1996) ("[Slection 10-4-609 regulates coverage for injuries caused by uninsured motorists and therefore governs the terms of the insurance contract.") (internal citation omitted).

III. The Use Test

The General Assembly has declared that the purpose of the No Fault Act is to avoid inadequate compensation to victims of automobile accidents; to require registrants of motor vehicles in this state to procure insurance covering legal liability arising out of ownership or use of such vehicles and also providing benefits ... to persons injured in accidents involving such vehicles. 10-4-702, 3 C.R.S. (2002). The Act requires insurers to pay certain expenses related to injuries "arising out of the use or operation of a motor vehicle." 10-4-706(1)(b)(D), 3 C.R.S. (2002). The UM statute, by contrast, does not absolutely mandate coverage but instead requires insurers to provide coverage against uninsured motorists for injuries "arising out of the ownership, maintenance or use of a motor vehicle," unless rejected in writing by the insured. 10-4-609(1)(a), 8 C.R.S. (2002). We have described the legislative intent underlying the UM Act as the General Assembly's desire "to provide compensation for injury caused by an uninsured motorist equal to that obtainable for injury caused by an insured motorist." State Farm Mut. Auto. Ins. Co. v. Nissen, 851 P.2d 165, 168 (Colo.1993).

At issue in this case is the meaning of the phrase "arising out of the use of a motor vehicle," found in both statutes. As discussed below, we have previously interpreted this phrase in many different settings. Because the language in each statute is virtually identical and because each statute derives from the same legislative purpose, our interpretation of the phrase "arising out of the use of a motor vehicle" applies to both. Compare Trinity Universal Ins. Co. v. Hall, 690 P.2d 227, 280 (Colo.1984) (construing the phrase "arising out of the use ... of a motor vehicle" for purposes of the No Fault Act) and McMichael, 906 P.2d at 102 (citing to Holl and construing the term "use" for purposes of the UM Act) and Kohli v. Union Ins. Co., 781 P.2d 184, 185-86 (Colo.1987) (construing the phrase "on account of the use of a motor vehicle" for purposes of Colorado's lability statute, § 42-7-102, 11 C.R.S. (2002)) with Cung La v. State Farm Auto. Ins., 880 P.2d 1007, 1009-12 (Colo.1992) (construing "arising out of the use" of a motor vehicle for purposes of the UM Act, noting that its meaning must be construed identically to the construction of "arising out of the use" for purposes of the Liability Act, also construing "arising out of the use" of a motor vehicle for purposes of the No Fault Act in the same *1261case, applying the same analysis for both the UM and No Fault provisions).

The terms "use" and "arising out of" are not statutorily defined terms for insurance purposes, and, as we have indicated above, they are not otherwise defined in the insurance contract. Thus, we must look to the intent of the parties at the time of contracting, taking into account any legislative intent that would impact upon the issue. See Nis-sen, 851 P.2d at 166 ("Our starting point is the plain language of the [insurance] contract and the intent of parties as expressed in that language.").

IV. Standard of Review

We review de novo the court of appeals' decision affirming the trial court's order of summary judgment in favor of the insured, Kastner. West Elk Ranch, LLC. v. U.S., 65 P.3d 479, 481 (Colo.2002). the appropriate standard of review whenever we consider a lower court's interpretation of our previous cases and Colorado statutes in a summary judgment context. Simpson v. Bijou Irrigation Co., 69 P.8d 50, 58 (Colo.2003). Of course, the factual findings here result from stipulated submissions of the parties, and thus it is the application of the law to those findings that we consider. This is

V. Analysis

1. Use of the Vehicle

As a threshold matter to recovery under either the uninsured motorist (UM) or personal injury protection (PIP) provisions of a policy, the claimant must show that at the time of the "accident," 1 the vehicle was being "used" in a manner "contemplated by the policy in question." Mason v. Celina Mut. Ins. Co., 161 Colo. 442, 444, 428 P.2d 24, 25 (Colo.1967) (quoting 7 Apple-man, Insurance Law and Practice § 4817)2 While the parties to an insurance contract may certainly contract for coverage beyond that required by law, Kastner and State Farm did not do so, and Kastner's policy does not define or expand upon "use." Accordingly, we apply the basic rules of insurance contract interpretation. See 7 Russ & Segalla, Couch on Insurance § 109:17 ("[It is to be presumed that the parties contracted with the intention of executing a policy satisfying the statutory requirements, and intended to make the contract to carry out its purpose."). Because we construe ambiguities in the insurance contract against the drafter, we will generally find coverage if the use in question is one that the insured claims to have contemplated or intended at the time of contracting for insurance. McMillan, 925 P.2d at 798; see also 8 Russ & Segalla, Couch on Insurance § 119:87 (explaining that a covered use "extends to any activity in utilizing the insured vehicle in the manner intended or contemplated by the insured"). Where the intent of the insured is not clear, McMichael instructs us to determine the covered "use" by looking at all the factual circumstances, "including the particular characteristics of the vehicle and the intention of [both] parties to the insurance contract." McMichael, 906 P.2d at 102. In all cases, however, the "use" in question must inhere in the nature of the insured automobile.

Our decision in McMichael clarified that "use" is a separate threshold inquiry. In McMichael, we explained that "[the first issue we must determine is whether [the claimant] was using an insured vehicle in a manner not foreign to its inherent purpose at the time of the accident." Id. at 101; see also 8 Russ & Segalla, Couch on Insurance §©119:37 ("[Tlhe concepts of use and legal cause should be analyzed separately, avoid*1262ing the traditional proximate cause concepts."). Whatever the use, it must be one that was contemplated by the parties to the insurance contract, and must be "inherent in the nature of the automobile [] as such." Mason, 161 Colo. at 444, 428 P.2d at 25 {internal citation omitted).

While "in general, operation of a motor vehicle for transportation purposes would constitute use, our cases demonstrate that use may have a broader meaning." McMichael, 906 P.2d at 102. In McMichael, we stated that to determine "use," "a court must look to the factual circumstances in each case." Id. Yet, even before McMichael, we clarified that "use" of a motor vehicle will include only those uses that are "conceivable" at the time of contracting for insurance and "not foreign to [the vehicle's] inherent purpose." Kohl, 781 P.2d at 186 n. 2.

Some vehicles may have an inherent non-transportation purpose that is plain and obvious to all contracting parties given the nature of the vehicle in question. Thus, McMichael directs us to look for factors that "adequately establish" whether the use in question was "conceivable and foreseeable at the time the parties entered the insurance contract." McMichael, 906 P.2d at 108. Our cases applying the term "use" before McMichael suggested that, unless the insurance contract provides otherwise, the only "conceivable use" not "foreign to [the] inherent purpose" of a non-commercial passenger vehicle is use as a means of transportation.

In Hall, we found use of the vehicle as a refreshment stand was "use" under the No Fault Act. At the time of contracting for insurance in that case, the vehicle was a factory-modified mobile refreshment stand. Fall, 690 P.2d at 231 n. 4, Accordingly, use of the vehicle to serve refreshments was clearly a "conceivable use" inherent in the vehicle's nature and thus both the insured and the insurer assumed the risks associated with that use. In Titan Construction Co. v. Nolf, 188 Colo. 188, 515 P.2d 1128 (Colo.1973), we determined that the unloading and loading of cement from a ready-mix cement truck constituted a "use" within the meaning of a Tiability-to-third-persons policy provision because such use was inherent in the nature of the vehicle. See Id. at 198-94, 515 P.2d at 1125-26. In McMichael, we concluded that a road construction worker who was sawing concrete barriers in the median of a highway some distance in front of his truck was "using" his vehicle as contemplated by the UM policy. McMichael, 906 P.2d at 108. There, the vehicle was a truck with factory-equipped overhead beacon and emergency flashers. The claimant was using the truck as a barricade and as a warning to cars on the highway when a car accidentally hit him. Id. at 94. Because the truck had emergency warning equipment in place at the time the parties entered into the insurance contract, we concluded that use of the vehicle for warning purposes was "conceivable and foreseeable" at the time of contracting. Id. at 108.

Although the term "use" is broad enough to cover activities beyond mere "transportation," it is not so broad as to include acts that are clearly independent of a vehicle's operation. See 1 No Fault and Uninsured Motorist Automobile Insurance § 9.10[2] (Matthew Bender 1984, updated 2008). In all cireumstances, the "use" in question should be one that is "foreseeably identifiable with normal use of a vehicle as a vehicle." 1 Irvin E. Schermer, Automobile Liability Insurance § 7:2[2] (8d ed.1995, updated 2003). While our past cases have not specifically equated "use" with transportation, in situations where the vehicle was a non-commercial passenger vehicle, the focus of the "use" inquiry has undeniably been on its connection to transportation. As these cases suggest, unless articulated otherwise in the policy, the only use of a non-commercial passenger vehicle that is foreseeable or conceivable at the time of contracting for insurance is use as a means of transportation.

In contrast to the commercial vehicles described above, a passenger vehicle has no obvious inherent use apart from its purpose as a mode of transportation. In Mason, we found that merely occupying or sitting in a parked car does not constitute a conceivable "use" for purposes of the insurance contract. Mason, 161 Colo. at 444, 428 P.2d at 24-25. By contrast, in Kokl, we held that transportation (including the loading and unloading) of hunters and their rifles was a conceivable *1263and foreseeable use at the time an insurance policy was signed for a four-wheel drive vehicle. Kohl, 781 P.2d at 136. Finally, in Cung La, we determined that because both the claimant's and assailant's vehicles were proceeding down a highway at the time the assailant shot the claimant, the cars were being "used" as motor vehicles. Cung La 830 P.2d at 1011.

The conclusion from our cases is, thus, that absent some plain and obvious special purpose, the normal use of the ordinary passenger car is limited to transportation.

Cases from other jurisdictions have articulated the transportation requirement more plainly. Requiring that at the time of the accident the vehicle was being used as a mode of transportation, these states give effect to the intent of their respective No Fault or UM statutes while recognizing the general purpose of contracting for automobile insurance.3 See, eg., McKenzie v. Auto Club Ins. Ass'n, 458 Mich. 214, 580 N.W.2d 424, 426 (1998) (Interpreting the statutory term "use . as a motor vehicle," the court held "we are convinced that the clear meaning of this part of the no fault act is that the Legislature intended coverage of injuries resulting from the use of motor vehicles when closely related to their transportational function and only when engaged in that function."); see also Ill. Farmers Ins. Co. v. League of Minn. Cities Ins. Trust, 617 N.W.2d 428, 429 (Minn.Ct.App.2000), (requiring that the vehicle being "used" must be "in the business of transporting persons or property"); see also Commercial Union Assurance Cos. v. Howard, 637 S.W.2d 647, 649 (Ky.1982) ("Basic automobile insurance policies are intended to cover 'driving' the vehicle, not repairing it. This additional field of coverage should be provided for by appropriate policies intended for that particular purpose.").

On the other hand, many courts are more liberal than Colorado on this point. Like Colorado prior to McMichael, those courts tend to collapse the "use" determination into the causal determination by asking whether the injury is related to the use of the car without first determining what the "use" is. See, e.g., Blish v. Atlanta Cas. Co., 736 So.2d 1151, 1155 (Fla.1999) (holding in favor of the insured and reasoning that the insured's attack by strangers along the highway while changing his tire related to the "use" of the vehicle since the attack was an "eminently foreseeable consequence of the use and maintenance of the" insured's truck).

2. Causal Connection between Use and Injury

If the use is foreseeably identifiable with the inherent purpose of a motor vehicle, the next prong of the inquiry is whether the "use" is causally related to the claimant's injury.

Beginning with our decision in Mason, we have consistently interpreted the phrase "arising out of the use" of a motor vehicle as requiring some causal connection between the "use" of the motor vehicle and the injury complained of, Mason, 161 Colo. at 448-44, 423 P.2d at 24-25. While occasionally referred to as a "but for" causal test, we have always required a claimant to show something more than a mere "but for" relation between the use of the vehicle and the injury. We have also required something less than proximate cause in the tort sense.

In Kohl, we stated that "tlo establish the requisite causal relationship, the claimant must establish that the accident would not have occurred but for the vehicle's use." Kohl, 731 P.2d at 185 (emphasis added). In that same opinion, however, we also reaffirmed our previous decision in Titan Construction Co. and emphasized that the " 'but for' doctrine should not apply when there is a lack of relationship between the [vehicle] and the accident." Id. at 186 (citing to Titan Constr. Co., 183 Colo. at 195, 515 P.2d at 1126). Titan Construction Co. explained that courts will not assess the causal relationship according to the realm of torts, but rather according to contract causation analy*1264sis. Titan Constr. Co., 183 Colo. at 194, 515 P.2d at 1126.

Several later cases restated the tight "but for" test first enunciated in Titan Construction Co. but also obscured it slightly by including language that first appeared in Azar v. Employers Cas. Co., 178 Colo. 58, 495 P.2d 554 (Colo.1972). In Azar, we interpreted "arising out of the use" of a vehicle only to require that the injury "originate from, grow out of, or flow from" the use of the vehicle. Azar, 178 Colo. at 61, 495 P.2d at 555 (internal quotations omitted). We also held in Azar that "there must be a causal relation or connection between the injury and the use of a vehicle in order for the injury to come within the meaning of the phrase 'arising out of the use' of a vehicle." Id. Titan Construction Co., decided one year after Azar, and interpreting a liability policy nearly identical to the one in Azar, made no mention of the test formulated in that case. Nevertheless, we have cited the language of both decisions in PIP, liability and even UM cases, often simultaneously.4 See, e.g., Hall, 690 P.2d at 231.

In Cung La, we stated that recovery depended not only upon the "nexus" between the motor vehicle and the injuries complained of, but also upon whether the injury would not have occurred "but for" the use of the vehicle and whether the injury flowed from or arose out of the use of the vehicle. Cung La, 880 P.2d at 1012.

Finally, our most recent case to interpret the phrase "arising out of the use" of a motor vehicle was McMichael. In McMichael, we described the causal analysis as requiring a claimant to "show that the accident would not have occurred but for the vehicle's use." McMichael, 906 P.2d at 108. There, we equated this phrase with a showing "that the injury originated in, grew out of, or flowed from a use of a vehicle." Id. We also held that the claimant must show that the vehicle's use was "integrally related to the claimant's activities and the injury at the time of the accident." Id. (emphasis added). McMi-chael was also careful to point out that the "nexus guarantees that the accident is within the kind of risks that the automobile insurance contract was meant to cover." Id.

We interpret this series of cases as requiring not only a "but for" connection between the "use" of the vehicle and the claimant's injury, but also an unbroken causal chain between that use and the injury.

Under this framework, the claimant must first show that except for the use of the vehicle, the accident or incident in question would never have taken place. Since Titan Construction Co., this requirement has been explicit and continually treated on an ad hoe basis. See, eg., Cung La, 830 P.2d at 1012 (noting that to survive a summary judgment motion, the claimant must only show a material question of fact exists as to the initial "but for" determination).

In addition, to complete and satisfy the causal analysis, the claimant must show that the "use" of the vehicle and the injury are directly related or inextricably linked so that no independent significant act or non-use of the vehicle interrupted the "but for" causal chain between the covered use of the vehicle and the injury.

Where the injury in question suffered by the insured is actually the result of an intentional act of another, this showing can be particularly difficult to make. However, our cases, as well as cases in a significant number of other states, illustrate that the intentional, even criminal act of another will not automatically preclude recovery. See Cung La, 880 P.2d at 1011-12; see also McMillan, 925 P.2d at 794 (holding that the intentional nature of a drive-by shooting did not preclude a finding that injuries were "caused by an accident" for purposes of UM coverage; "[Tlhe intentional conduct of the uninsured tortfeasor [in Nissen ] did not preclude our holding that uninsured motorist coverage existed for the insured.").

*1265Instead, these cases demonstrate that the claimant may recover provided that the injury flows directly from the "use" of the vehicle, without interruption, so that the "use" of the vehicle and the resulting injury constitute "one ongoing assault." We have previously observed that using a car merely to help carry out a criminal act is not the kind of "risk that the automobile insurance contract was meant to cover," McMichael, 906 P.2d at 103. Instead, where the act causing the injury is intentional, the "use" of the vehicle must bear a direct relation to the assault.

For example, in Cung La, the insured was driving his own insured vehicle down the highway at the time his assailants shot him. Cung La, 830 P.2d at 1008. The assailants, too, were driving down the highway and used their uninsured vehicles to position themselves in order to shoot the insured. Id. We assumed that the assailants and the insured were "using" their cars as contemplated by the insured's policy since the cars were moving at the time of the shooting.

The issue, therefore, in Cung Lo was whether a jury could find, based on the evidence, that the shooting injuries arose out of this "use." Id. at 1011-12. In response to that issue, we determined as an initial matter that the shooting would not have occurred "but for" the victim's "use" of the car (he was identified by his assailants only by the car he was driving and thus would not have been shot except for the fact that he was using his ear as transportation). Id. at 1012. Because the covered "use" of the cars served as an active accessory to the shooting, and because there was no interruption or other independent significant act between this "use" and the shooting, a jury could have found that the injuries were sufficiently causally related. The act of driving and the act of shooting were inextricably linked with no intervening act. See Cont'l W. Ins. Co. v. King, 415 N.W.2d 876, 878 (Minn.1987)(holding there was no act of independent significance to break the causal link between the uninsured's shooting the insured while both were driving on a road). The result reached in Cung Lo is in line with the reasoning in cases from other jurisdictions with practically identical facts. See, eg., Wausau Underwriter's Ins. Co. v. Howser, 309 S.C. 269, 422 S.E.2d 106, 109 (1992) (per curiam); see also AIG Howair Ins. Co., Inc. v. Caraang, 74 Haw. 620, 851 P.2d 321, 330-31 (1998).

In Nissen, we permitted recovery where the uninsured hit the claimant with her own car, throwing her onto the hood and pinning her between two cars after driving into traffic. Nissen, 851 P.2d at 166. Because the uninsured was driving the car when he hit the claimant, we did not question whether the car was being "used" at the time of the assault. Additionally, because the car physically contacted the claimant, we did not question the causal relationship between the "use" and the injury in that case. Instead, resolution of Nissen turned on the interpretation of two conflicting terms in the insured's policy. Id.

VI. Application of the Two-Prong Test

Kastner's car was an ordinary noncommercial passenger car with no plain and obvious inherent purpose as a vehicle other than the safe transportation of its passengers and cargo. That determination shapes the balance of our analysis.

We conclude, unlike the court of appeals, that the "use of the reclining passenger seat to prevent [Kastner] from signaling for help;" "the use of the vehicle to get to an isolated area" to commit a crime; and the "use of the automatic seat belts as restraints," Kastner, 56 P.3d at 1146, were all "foreign to the inherent purpose" of the motor vehicle as a mode of transportation. These uses, whether viewed individually or collectively, are not "uses" as contemplated by both the statutes and insurance policies in question. Use of a reclining passenger seat to conceal a kidnapping has little to do with using a car for transportation purposes. Use of a car to get to an isolated area to commit a crime may relate to a vehicle's general transportation purpose, but here it was not concurrent with the injury itself, and, as explained below, it lacks the requisite causal connection between sexual assault and "use" of a car for transportation. Finally, use of the car's seatbelts to restrain a sexual assault victim relates neither to the vehicle's transportation purpose nor to any other "conceiva*1266ble" or foreseeable use contemplated at the time of contracting for insurance.

None of these uses-individually or collectively-comprise "use" of a passenger vehicle in a manner foreseeably identifiable with the ordinary use of that vehicle. In the case before us today, the most we can say about the assailant's use of the car was that it served as the site of the sexual assault and that the assailant employed the car's furnishings to help complete the assault inside the car. These uses are not foreseeably identifiable with the inherent purpose of a motor vehicle.

Additionally, the facts do not establish a causal nexus between a covered "use" of the vehicle and the victim's injuries. Thus, Kast-ner's claim also fails the second prong of this analysis. The seat belt and the reclining seat served as accessories to the crime, merely assisting the assailant in a way that incidental objects or furnishings inside a house could have helped him without actually causing the assault. See Am. Nat'l Prop. & Cas. Co. v. Julie R., 76 Cal.App.4th 184, 142, 90 Cal.Rptr.2d 119 (1999). Similarly, using the car to drive the victim to a remote location no more connects the car to the assault than if the assailant had used the car as the mere situs of the assault without moving it. See id. at 140, 90 Cal.Rptr.2d 119; see also Sanchez v. State Farm Mut. Auto. Ins. Co., 878 P.2d 31, 33 (Colo.App.1994)("[TJhe mere transportation of the dog [that bit the claimant] to the scene of the injury is, by itself, insufficient to support a finding that the injury arose from the use of the automobile."); see also Klug, 415 N.W.2d at 878.

Kastner argues that her own use of the car further connects the assault to the vehicle. Specifically, she asserts that her opening of the car door made the ensuing trapping and assault possible, establishing the "but for" connection. She states that the assailant had not chosen her as a victim until she opened her car door. Even assuming this constitutes a "use" of the car, the relationship between this use and the resulting sexual assault is simply too tenuous. All of the other non-uses of the car interrupted any direct flow between this "use" and the injury. Accordingly, the use of the car and the injury were not causally linked so as to make the use of the car and the injury one ongoing assault.

The approach recommended by Kastner raises additional concerns. She argues that each insurance claim that somehow involves both a motor vehicle and a sexual assault should be reviewed individually, letting the jury decide whether in each case the causal nexus between vehicle and injury justifies recovery. Kastner concedes that each of the "uses" of her car in this incident on its own may not constitute a sufficient relationship between injury and car. But, collectively, she asserts that the four facts of the case justify recovery. As amici have noted, this case-by-case approach would inevitably lead to inconsistent and unfair results. For instance, under Kastner's recommendation, she would recover because the four facts-the opening of the car door, the seat belt, the reclining seat, and the driving of the car-together satisfy the causal nexus. Yet, the next victim would not recover because her assailant used his own restraints instead of the seat belt. Or, another victim would not recover because, though the assailant used the seat belt, the reclining seat, and the opening of the car to identify his victim, the assailant did not drive the car anywhere. Both the failing and succeeding combinations of facts are endless, and the point is clear. Every case of sexual assault somehow involving a car would go before a jury and, arbitrarily, some victims would recover and some would not.

VII. Conclusion

In conclusion, we hold that Kastner cannot recover under her UM and PIP policies for injuries related to the sexual assault because (1) at the time of the accident, the vehicle was not being "used" in a manner that was reasonably foreseeable at the time of contracting for the policies and (2) Kastner's injuries had an insufficient causal nexus with the use of the vehicle. Accordingly, we reverse the decision of the court of appeals and return the case to the trial court with directions to enter summary judgment on State Farm's motion.

*1267Justice BENDER dissents and Chief Justice MULLARKEY and Justice MARTINEZ join in the dissent.

Justice BENDER,

dissenting:

In my view, the majority's test is flawed in two ways. First, our Colorado cases do not require concurrency between use and injury. Second, the majority's causation prong misreads our prior holdings. According to my reading of the cases, the test should be whether an injury originates in, grows out of, or flows from the use of a car. Applying either the majority's test or my suggested test, Christina Kastner should recover because her ongoing assault, kidnapping, and rape were causally related to the transportation use of her car.

Admittedly, ambiguities exist in our accident recovery cases. The majority's attempt to synthesize these cases is no enviable task, and I do not fault the majority for seeking to achieve clarity from these disparate and somewhat inconsistent holdings. However, I read our precedent differently and would apply different rules to determine whether our statutes and Kastner's insurance policy permit her recovery. Therefore, I respectfully dissent.

L.

Unlike general contract terms, "the provisions in [an insurance] policy are often imposed on a take-it-or-leave-it basis. It is not a negotiated contract but one with terms required by legislation or dictated by an insurer." Huizar v. Allstate Ins. Co., 952 P.2d 342, 344 (Colo.1998). Accordingly, this Court "assumes a "heightened responsibility' to serutinize" insurance policy provisions to ensure that they comply with "public policy and principles of fairness." Id. Further, the legislative history of the uninsured motorist statute "instructs us to find coverage for the innocent insureds whenever possible." State Farm Mut. Auto. Ins. Co. v. Nissen, 851 P.2d 165, 169 (Colo.19983). For this reason, ambiguities in insurance coverage should be construed against the insurer. Id. at 166.

A.

As the majority points out, the foreseeable use of a car is generally limited to its transportation purpose.1 See maj. op. at p. 1262-1263. However, the majority requires that Kastner's injuries be concurrent with her vehicle's use. See maj. op. at p. 1265 ("Use of a car to get to an isolated area to commit a crime may relate to a vehicle's general transportation purpose, but here it was not concurrent with the injury itself ...." (emphasis added)). The majority appears to derive this concurrency requirement from a sentence in Aetna Cas. & Sur. Co. v. McMichael, 906 P.2d 92, 101 (Colo.1995) ("The first issue we must determine is whether [the insured] was using an insured vehicle in a manner that was not foreign to its inherent purpose at the time of the accident.")(emphasis added). As a preliminary matter, the majority's concurrency requirement contradicts our traditional causation test, under which an injury need only originate in, grow *1268out of, or flow from the use of a vehicle, regardless of when it occurs.2 More importantly, though, we have never required concurrency between use and injury. For instance, Kohl v. Union Ins. Co., 731 P.2d 134 (Colo.1987), involved a group of hunters on their way home from a hunting trip who stopped briefly at a convenience store. While the group was conversing in the store's parking lot, one of the hunters decided to unload his rifle in a jeep when it accidentally discharged, killing a fellow hunter and seriously injuring two others. Id. at 185. This Court held that the injuries, though not concurrent with the jeep's use at the time of the injuries, were covered under the insured's policy. Id. at 185-36. Contrary to what the majority suggests, McMi-chael did not change this holding.

B.

In its causation prong, the majority imposes a strict but-for analysis and borrows its "independent significant act" element from other jurisdictions.

My first criticism of the majority's causation prong is its misplaced reliance on a strict but-for test. Specifically, the majority misreads McMichael as requiring, at a minimum, but-for causation. See maj. op. at p. 1263 ("[Wle have always required a claimant to show something more than a mere 'but for' relation between the use of the vehicle and the injury."). To justify its strict but-for test, the majority focuses on a single sentence from McMichael in which we described the evolution of our causation analysis. See maj. op. at p. 1264. The sentence explained that in KoRl, we used a but-for test as a threshold causation requirement. However, the majority fails to mention that the very next sentence of McMichael significantly qualifies the causation analysis from Kohl by stating that our precedent follows a "more liberal interpretation" requiring "only that the injury originated in, grew out of, or flowed from a use of a vehicle." 906 P.2d at 103. Thus, as I read McMichael, we specifically rejected the strict but-for test:

In Kohl, we explained that in order to establish the requisite causal relationship between the use of the vehicle and the injury, the claimant must show that the accident would not have occurred but for the vehicle's use. Although the use of "but for" terminology suggests that the use of the vehicle must be the cause of the injuries, we have utilized a more liberal interpretation in our cases. . .. In fact, we have interpreted the test as requiring the plaintiff to show only that the injury originated in, grew out of, or flowed from a use of a vehicle.

Id. (emphasis added) (citations omitted).

My reading of McMichael as requiring something less than but-for causation finds support in the statute and our previous decisions. As the majority notes, the insurance statutes require coverage for injuries "arising out of the use of a motor vehicle." Since our 1972 decision in Azar v. Employers Cas. Co., 178 Colo. 58, 495 P.2d 554 (1972), we have "broadly and comprehensively" construed the "arising out of" language to mean " 'originate from,' 'grow out of, or 'flow from.' " Id. at 61, 495 P.2d 554. Decades later, McMichael confirmed that our cases require something less than but-for causa tion. 906 P.2d at 108 (Vollack, C.J., dissenting)(noting that the insured did not satisfy the but-for test).

Our cases have consistently interpreted the "originated in, grew out of, or flowed from" language in a flexible way. In Azar, we interpreted the phrase as requiring only that a vehicle "contribute[ ] to or [be] connected to" an injury. 495 P.2d at 555. We have used equally flexible terms in our most recent decisions. See McMichael, 906 P.2d at 1083-104 ("related to"); Cung La v. State Farm Auto. Ins. Co., 880 P.2d 1007, 1010 (Colo.1992) ("contributed to"). Thus, our causation test has traditionally been satisfied where a vehicle's use contributed to, or was connected or related to, the insured's injury. Thus, the majority's holding that a vehicle's use must be "inextricably related" to an insured's injury appears to contradict our case law.3

*1269My second criticism of the majority's causation prong is that it borrows the "independent significant act" element of its causation test from other jurisdictions,4 and the ele ment finds little or no support in our cases. In our cases, an intervening act by itself does not break the causal chain between use and injury. See Cung La, 830 P.2d at 1011 ("Here, the fact that the firearm contributed to the injuries does not preclude the requisite causal connection."). Thus, intervening acts, even intentional criminal acts such as the shootings in Cung La and State Farm Mut. Auto. Ins. Co. v. McMillan, 925 P.2d 785 (Colo.1996), and the auto theft in Nissen, do not preclude recovery where a vehicle's use contributes to the injury-causing acts. See Cung La, 830 P.2d at 1011; Nissen, 851 P.2d 165. For example, the insured in Nissen looked out the window of the restaurant where she was eating and saw a thief entering and attempting to steal her car from the restaurant parking lot. She ran outside, jumped on the hood of the car, and was injured when the thief attempted to flee while she remained spread-eagle on the car's hood. In this case, we approved the insured's recovery under her uninsured motorist policy for the severe injuries she suffered when the thief crashed her car into an oncoming truck. 851 P.2d at 165.

Instead of holding that independent significant acts bar recovery, we have consistently concluded that an insured may not recover where a vehicle serves as the "mere situs" for an injury. McMichael, 906 P.2d at 108-104. Our cases have narrowly construed the mere situs restriction to mean that a vehicle is the mere situs for an injury if it in no way contributes to it. See id. at 104 (discussing Mason v. Celina Mut. Ins. Co., 161 Colo. 442, 423 P.2d 24 (Colo.1967), in which a youth was accidentally shot when he and two friends were toying with a pistol in a parked car and "the event could have occurred on the street, in a house, or on a porch"). In Kohl, where the hunter unloaded his rifle near the gun rack inside the jeep, we found that the injury was "intimately related" to the jeep's use even though the gun could have been unloaded outside the jeep. Id. Similarly, in both Cung Lo and McMillan the shooters could have waited until their respective victims got out of their cars to shoot them, but because the cars contributed to the shootings, we found that they were not the mere situs of the crimes. See Cung La, 830 P.2d 1007; McMillan, 925 P.2d 785. In short, the mere situs restriction only applies when the vehicle does not contribute to the injury in any way.

Based on these cases, I believe the proper test for causation should be whether an injury "originated in, grew out of, or flowed from" the use of a vehicle This test is satisfied where a vehicle's use contributes to an injury unless the vehicle was the "mere situs" of the injury.

IL

Irrespective of whether the majority's view or my view on the issues of use and causation is correct, Kastner should recover.

I agree with the majority's statement of the facts with one addition: Kastner's ordeal, which began in the mall parking lot, took place during the Christmas shopping season.5 This fact helps to explain why the mall parking lot was full at the time of Kastner's abduction, and why it is highly unlikely that Kastner's assailant could have kidnapped or raped her without the use of her car.

My primary objection to the majority's analysis is that it focuses exclusively on Kast-*1270ner's rape injuries, all of which occurred while the car was stopped, without considering the ongoing kidnapping and assault violations. Without question, Kastner's injuries were ongoing. They began in the mall parking lot, where she was assaulted with a knife and kidnapped. While driving Kastner to the park, the assailant continued to hold the knife on her as he threatened to kill her. During the brief period when he stopped the car in the park, the assailant raped and robbed Kastner. On the way to the liquor store where he dropped her off, the assailant threatened severe harm to Kastner and her children if she reported the assault. Kastner suffered harm throughout her ordeal, not just during the rape.

A.

Even applying the majority's test, Kastner should recover. Because the majority focuses exclusively on the injuries caused by the rape, it concludes that Kastner's injuries were not concurrent with the car's use. However, Kastner's ongoing injuries-the kidnapping and extended assault-were concurrent with the vehicle's use because they both occurred while Kastner's car was being used for transportation purposes. Moreover, the majority concedes that the use of Kast-ner's car to get to the park relates to the car's general transportation purpose. See maj. op. at p. 1265.

Although the majority concludes that Kast-ner fails to satisfy the but-for and "independent significant act" tests under its causation prong, the facts of this case undermine this conelusion. But for the use of Kastner's vehicle, it is highly unlikely that the assailant would have been able to kidnap her from a crowded parking lot during the Christmas shopping season. It is also highly unlikely that Kastner would have been raped if the assailant could not have used her car to transport her from the mall. Kastner's assailant escaped detection by driving Kastner to a secluded park on a dark winter night. It is also highly unlikely that the rape would have occurred in a busy mall parking lot during Christmas season in a car parked only ten spots from the mall entrance. Thus, I conclude that but for the use of the car, Kastner would neither have been raped nor kidnapped. In any case, causation findings are within the province of the trier of fact, and the trial court specifically found that Kastner's injuries arose from the use of her car.

Under the rationale of our Colorado cases, independent significant acts like the assailant's rape and kidnapping of Kastner do not break the causal chain between use and injury. We have not regarded comparable intentional criminal conduct, like the shootings in McMillan and Cung Lo and the auto theft in Nissen, to constitute independent significant acts sufficient to defeat causation. Similarly, the criminal acts of Kastner's assailant did not break the chain of causation. Thus, Kastner satisfies the majority's strict tests for use and causation and she should be entitled to recover.

B.

Applying the test supported by my reading of our cases, Kastner should also recover. Beyond doubt, Kastner's kidnapping and ongoing assault were sufficiently temporally related to the transportation use of her car. The fact that Kastner's assailant stopped her car briefly to rape her should not prevent her from recovering. Kastner's injuries during this brief stop are analogous to the accidental shooting in Kohl, which occurred while the hunters' jeep was parked. We allowed the insured to recover in that case, and we should in this case as well. As in Kokl, the injuries that occurred while the car was stopped in this case are sufficiently related to the car's transportation use to warrant recovery.

Moreover, Kastner's injuries were causally related to her car's use. The assailant's use of Kastner's vehicle contributed both to Kast-ner's kidnapping and rape. Kastner's assailant was able to kidnap her from a crowded mall during the Christmas shopping season only by transporting her in the car. The rape was facilitated by the assailant's ability to remove Kastner to a remote area. Kast-*1271ner's car was not the mere situs of her injuries. The car contributed to Kastner's kidnapping and rape in at least as significant a way as the vehicles in McMillan and Cung La contributed to the shootings in those cases. Therefore, I respectfully dissent.

I am authorized to state that Chief Justice MULLARKEY and Justice MARTINEZ join in this dissent.