FOR PUBLICATION

UNITED STATES COURT OF APPEALS

FOR THE NINTH CIRCUIT

ESTATE OF DECEDENT No. 23-4466

LOLOMANIA SOAKAI; LAVINIA

D.C. No. SOAKAI, an individual and personal 3:23-cv-00381-SK representative of Estate; DANIEL FIFITA, an individual; SAMIEUELA FINAU, an individual; INA OPINION LAVALU, an individual,

Plaintiffs - Appellees,

v.

WALID ABDELAZIZ, in his individual capacity as a police officer for the city of Oakland; JIMMY MARIN-CORONEL, in his individual capacity as a police officer for the city of Oakland,

Defendants - Appellants,

and

CITY OF OAKLAND, a municipal corporation,

Defendant. Appeal from the United States District Court for the Northern District of California Sallie Kim, Magistrate Judge, Presiding *

Argued and Submitted November 20, 2024 San Jose, California

Filed May 16, 2025

Before: Susan P. Graber, Michelle T. Friedland, and Patrick J. Bumatay, Circuit Judges.

Opinion by Judge Graber; Dissent by Judge Bumatay

SUMMARY **

Qualified Immunity

The panel affirmed the district court’s denial of defendant police officers’ motion for judgment on the pleadings based on qualified immunity in a 42 U.S.C. § 1983 action brought by innocent bystanders, who were injured by a fleeing suspect who lost control of his car and crashed into them as a result of a high-speed car chase.

* The parties consented to proceed before a magistrate judge. See 28 U.S.C. § 636(c). ** This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader.

ESTATE OF SOAKAI V. ABDELAZIZ 3

Plaintiffs alleged that defendants violated their Fourteenth Amendment substantive due process rights by (1) conducting a high-speed chase for the purpose of harming the fleeing suspect in a manner that exceeded any legitimate law enforcement purpose, and (2) failing to summon or render emergency services for plaintiffs after the crash that defendants affirmatively helped to cause. Addressing plaintiffs’ purpose-to-harm claim, the panel held that plaintiffs stated a substantive due process claim by plausibly alleging that, as bystanders, they were injured when defendants conducted a high-speed chase with a purpose to harm the suspect in a manner that exceeded any legitimate law enforcement purpose. Because the law was clearly established before the date of the car chase that defendants’ conduct was unconstitutional, defendants were not entitled to qualified immunity. The panel rejected defendants’ assertion that to state a Fourteenth Amendment substantive due process claim, a bystander injured by a high- speed police chase must plausibly allege that the officer acted with an improper purpose to harm the bystander specifically. This Circuit’s precedent recognizes that an officer owes a duty to all those in the vicinity, including bystanders, to limit their intent to harm to legitimate law enforcement purposes. Addressing plaintiffs’ alternative, narrower state-created danger claim, the panel held that—although the Fourteenth Amendment generally does not confer any affirmative right to governmental aid—plaintiffs plausibly alleged that defendants affirmatively created danger by initiating a car chase that led to a crash and then acted with deliberate indifference to plaintiffs’ worsening medical condition by failing to summon help. If plaintiffs’ allegations are true, defendants violated clearly established law by acting with deliberate indifference to the injuries that resulted from the collision that defendants affirmatively helped to cause. Dissenting, Judge Bumatay wrote that the officers were entitled to qualified immunity. The majority adopted a brand-new theory of substantive due process—contrary to precedent and to the Supreme Court’s admonition against such judicial overreach—by ruling for the first time that a bystander may assert a substantive due process claim against an officer if the bystander can show that the officer intended to harm someone else. Given that this novel theory of due process conflicts with Supreme Court and Ninth Circuit precedent, the law was not clearly established at the time of the accident that intent to harm a suspect is enough to press a due process claim for injuries to bystanders. The majority also expanded the state-created-danger doctrine to create a new constitutional duty requiring law enforcement officers to render or summon medical aid for civilians harmed by private actors under certain circumstances.

COUNSEL

Patrick M. Buelna (argued), Matthew Norman, and Lateef H. Gray, Pointer & Buelna LLP, Oakland, California; Adante D. Pointer, Burris Nisenbaum Curry & Lacy LLP, Oakland, California; for Plaintiffs-Appellees. David B. Newdorf (argued), Newdorf Legal, Oakland, California; Richard W. Osman and Sheila D. Crawford, Bertrand Fox Elliot Osman & Wenzel, San Francisco, California; Aimee G. Hamoy, Kaufman Dolowich & Voluck LLP, Oakland, California; for Defendants-Appellants.

ESTATE OF SOAKAI V. ABDELAZIZ 5

OPINION

GRABER, Circuit Judge:

Plaintiffs, a group of innocent bystanders, were injured by a driver who lost control of his car and crashed into them as a result of a high-speed car chase. Plaintiffs sued two police officers, claiming that the officers violated Plaintiffs’ substantive due process rights in two ways: by initiating and conducting the chase for the purpose of harming the fleeing suspect and by failing either to call for emergency services or to render aid after the crash. Assessing only the pleadings, the district court ruled that the officers are not entitled to qualified immunity. The officers timely appeal. In the highly unusual circumstances of this case—including plausible allegations that the officers intentionally caused harm for reasons unrelated to any legitimate law enforcement purpose connected to the chase, and that they witnessed the crash yet drove away and later stated that they hoped that the crash caused a fatality—we affirm.

BACKGROUND

The facts alleged in this case are jarring and tragic. We must take all plausible allegations as true at this stage of the proceeding. Al Saud v. Days, 50 F.4th 705, 709 (9th Cir. 2022). On June 25, 2022, in Oakland, California, Officers Jimmy Marin-Coronel and Walid Abdelaziz, police officers of the Oakland Police Department and Defendants in this action, spotted a person who, they believed, had participated in an illegal car rally. Even though the Oakland Police Department’s policy authorized high-speed car chases only in cases involving certain violent crimes, Defendants began pursuing the suspect through busy city streets at speeds exceeding 60 miles per hour. Allegedly intent on making the suspect crash, Defendants did not turn on their lights or sirens, nor did they report the chase to the dispatcher. Those actions, too, violated departmental policy. The chase ended when the suspect’s car smashed into an area near a popular taco truck, where Lolomania Soakai (“Lolomania”) had stopped with his family and friends on the way home from a graduation ceremony. Lolomania suffered a direct hit and died of his injuries in front of his mother, Plaintiff Lavinia Soakai (“Lavinia”), who broke her back in the crash. Other members of their group, including Plaintiffs Daniel Fifita, Ina Lavalu, and Samiuela Finau, also suffered severe injuries. Despite witnessing the crash, Defendants neither stopped to render aid nor summoned emergency services. Instead, Defendants drove by the scene—still with their lights and sirens off—and did not return until they heard other officers approaching the area of the crash. When they did return, Defendants pretended not to have been at the scene previously. While still at the site of the crash, Defendants were overheard saying that “they were satisfied the [suspect] appeared injured and hoped that the [suspect] had died in the crash.” Lolomania’s estate and the individual Plaintiffs sued, asserting claims under 42 U.S.C. § 1983 for violations of the Fourteenth Amendment. 1 Defendants moved for judgment

1 Plaintiffs also brought a claim for liability against the City of Oakland under Monell v. Department of Social Services, 436 U.S. 658 (1978), and a claim for violations of California’s Bane Act against the individual Defendants and the City. Those claims, which the district court dismissed, are not presently before us.

ESTATE OF SOAKAI V. ABDELAZIZ 7

on the pleadings, arguing that they were entitled to qualified immunity. The district court denied the motion with respect to Plaintiffs’ § 1983 claim, and Defendants filed this timely interlocutory appeal.

STANDARD OF REVIEW

We review de novo the denial of a motion for judgment on the pleadings premised on qualified immunity, Carrillo v. County of Los Angeles, 798 F.3d 1210, 1218 (9th Cir. 2015), employing the same standards used when reviewing a motion to dismiss, Gregg v. Haw., Dep’t of Pub. Safety, 870 F.3d 883, 887 (9th Cir. 2017). Consequently, in evaluating the parties’ arguments, “[w]e view the allegations in the complaint as true and in the light most favorable to [Plaintiffs].” Al Saud, 50 F.4th at 709.

DISCUSSION

Plaintiffs claim that Defendants violated their constitutional rights by (A) conducting a high-speed chase for the purpose of harming a fleeing suspect unrelated to any legitimate law enforcement purpose and (B) failing to summon or render emergency services after the crash. Both of those theories sound in the Fourteenth Amendment’s guarantee of substantive due process, which protects against “executive abuse[s] of power” that “shock[] the conscience.” County of Sacramento v. Lewis, 523 U.S. 833, 846 (1998). 2 For each theory, which we will address in turn, Defendants

2 In addition to representing her son’s estate, Lavinia seeks to recover for two different violations of her own due-process rights. The first stems from the injuries she personally suffered, and the other arises from the loss of a familial relationship. Because the same shocks-the-conscience standard applies to both types of claims, Porter v. Osborn, 546 F.3d 1131, 1137 (9th Cir. 2008), we will not differentiate between them in this opinion. are entitled to qualified immunity if they can show (1) that the allegations in the operative complaint, accepted as true, “do not make out a violation of a constitutional right”; or (2) “that any such right was not clearly established at the time of the alleged misconduct.” Hampton v. California, 83 F.4th 754, 765 (9th Cir. 2023). A. Purpose-to-Harm Claim Plaintiffs first contend that the complaint states a claim that Defendants violated clearly established law by chasing the fleeing suspect for an improper purpose, harming bystanders in the process. We agree. 1. Constitutional Violation a. Legal Framework To violate the substantive component of the Due Process Clause, official conduct must “shock[] the conscience.” Porter, 546 F.3d at 1137 (quoting Lewis, 523 U.S. at 846). Because behavior “that shocks in one environment may not be so patently egregious in another,” Lewis, 523 U.S. at 850, courts have developed two methods for defining conscience- shocking conduct: the deliberate-indifference test and the purpose-to-harm test, Scott v. Smith, 109 F.4th 1215, 1228 (9th Cir. 2024). We decide which test to apply by “ask[ing] ‘whether the circumstances are such that actual deliberation [by the officer] is practical.’” Wilkinson v. Torres, 610 F.3d 546, 554 (9th Cir. 2010) (second alteration in original) (quoting Porter, 546 F.3d at 1137). Because officers engaged in a high-speed chase must “operate under great pressure and make repeated split- second decisions” with “precious little time for deliberation,” Bingue v. Prunchak, 512 F.3d 1169, 1176 (9th Cir. 2008), we apply the more stringent purpose-to-harm test

ESTATE OF SOAKAI V. ABDELAZIZ 9

“to all high-speed chases,” id. at 1177 (emphasis omitted). Under that test, a police officer violates substantive due process only if the officer “act[s] with the purpose to harm a civilian” for reasons “unrelated to the legitimate law enforcement objectives of arrest, self-defense, or the defense of others.” A.D. v. Cal. Highway Patrol, 712 F.3d 446, 454 (9th Cir. 2013). Satisfying the purpose-to-harm test is inherently difficult for those injured by high-speed chases. Officers almost always chase a suspect for a legitimate law enforcement purpose and, even when they might not, legitimate justifications are readily available. See Porter, 546 F.3d at 1137 (explaining that officers chasing a suspect are generally “reacting to the urgent public safety threat of fleeing motorists in a situation where inaction could be the most dangerous option”); Lewis, 523 U.S. at 853 (noting that police pursuits can serve “the need to stop a suspect and show that flight from the law is no way to freedom”); cf. Scott v. Harris, 550 U.S. 372, 383 (2007) (stating, in a Fourth Amendment case, that the fleeing suspect had “posed an actual and imminent threat to the lives of any pedestrians who might have been present, to other civilian motorists, and to the officers involved in the chase”). Indeed, we are unaware of any car-chase case in this circuit in which a plaintiff passed the purpose-to-harm test on the merits. See Onossian v. Block, 175 F.3d 1169, 1172 (9th Cir. 1999) (affirming grant of summary judgment to officers because the evidence suggested that “they were attempting to remove a dangerous driver from the streets,” not to cause harm); Bingue, 512 F.3d at 1177 (similar). The task is not impossible, however, because the purpose-to-harm test turns on the officer’s subjective intent. See A.D., 712 F.3d at 453 (“The purpose to harm standard is a subjective standard of culpability.”). A court employing that test will not second-guess an officer’s decision—even if seemingly ill-advised in hindsight—if the officer “acts with a legitimate purpose . . . in mind.” Tan Lam v. City of Los Banos, 976 F.3d 986, 1003–04 (9th Cir. 2020). But the availability of an otherwise plausible excuse will not shield the officer from liability if the officer acts with an “ulterior motive[]” to harm that is unrelated to legitimate law enforcement purposes. Gonzalez v. City of Anaheim, 747 F.3d 789, 798 (9th Cir. 2014) (en banc); see also A.D., 712 F.3d at 453 (noting that, even if an officer ultimately arrests the suspect, “he still violates the [D]ue [P]rocess [C]lause if he used force with only an illegitimate purpose in mind”). At the motion-to-dismiss stage, the key question is whether, accepting all well-pleaded allegations of fact as true, the officer subjectively intended to act, not to further a legitimate law enforcement purpose, but instead to induce lawlessness, to terrorize, to cause harm, to kill, to teach the suspect a lesson, or to get even. Porter, 546 F.3d at 1140– 41; see also Zion v. County of Orange, 874 F.3d 1072, 1077 (9th Cir. 2017) (holding that an officer could be liable for a physical assault if he “was acting out of anger or emotion rather than [to achieve] any legitimate law enforcement purpose”). b. Analysis Here, Defendants concede that the complaint adequately pleaded that they acted with a purpose to harm the fleeing suspect for reasons unrelated to a legitimate law enforcement objective. The complaint alleges that Defendants acted “in an effort to make [the] suspect lose control, severely injure himself[,] and die.” Defendants admit that we must treat as true that allegation, see Opening Br. at 21 n.2 (“[T]he Court

ESTATE OF SOAKAI V. ABDELAZIZ 11

must accept as true the allegation of the [complaint] that [D]efendants had intended to harm the suspect.”), the plausibility of which Defendants do not challenge. 3 The complaint also alleges that “Defendants use[d] their law enforcement powers to cause unnecessary harm to a person,” (emphasis added), and Defendants do not clearly and distinctly argue otherwise. To the contrary, Defendants premise their argument on the assumption that Plaintiffs pleaded a purpose to harm the suspect unrelated to a legitimate law enforcement objective. See id. at 21 (“If the officer has evil intent that shocks the conscious [sic] when it comes to the suspect, does that matter if the plaintiffs are the bystanders? That is the key question in the present case.” (emphasis added)). Defendants’ sole contention on appeal is that any improper intent to harm the fleeing suspect is irrelevant here because, in the context of a purpose-to-harm claim, the object of an officer’s improper intent to harm and the injured plaintiff must be one and the same. Put another way, Defendants assert that, to state a claim, a bystander injured by a high-speed police chase must plausibly allege that the officer acted with an improper purpose to harm the bystander specifically. We reject Defendants’ argument. In Lewis, the Supreme Court applied “a much higher standard of fault than deliberate indifference,” Lewis, 523 U.S. at 852, because an officer deciding whether to give chase must quickly balance the benefits of pursuit with the risks involved, id. at 853. When addressing the benefits, the Supreme Court

3 Defendants did not raise the issue of plausibility in their opening brief. And at oral argument, Defendants’ lawyer confirmed that Defendants were not contesting the plausibility of Plaintiffs’ allegation. concentrated on suspects, citing “the need to stop a suspect and show that flight from the law is no way to freedom.” Id. That focus makes sense; an officer chasing a suspect is, presumably, typically motivated by something related to that suspect. When discussing the risks, though, Lewis took a broader view, noting “the high-speed threat to all those within stopping range,” including “suspects, their passengers, other drivers,” and “bystanders.” Id. (emphasis added). Implicit in Lewis’s discussion is the recognition that, although the object of a high-speed police chase might be to catch the fleeing motorist, an officer owes a duty to all those in the vicinity, including bystanders. See Onossian, 175 F.3d at 1171 (noting that, in Lewis, “the duty of the pursuing police officer is defined generally, without specific reference to the suspect being pursued”). Lewis thus confirms what common sense dictates: High- speed car chases create a clear, known risk of harm, not only to the fleeing driver and to the officers, but also to passengers and bystanders. Because the risks taken by those participating in the chase generate—and, thus, cannot be isolated from—the peril faced by bystanders, it would be illogical to distinguish between those dangers when considering whether an officer ought to be liable for injuries that result from the decision to give chase. Lewis’s application of the purpose-to-harm test solidifies that principle. Lewis involved an officer’s pursuit of a motorcycle carrying two teenagers—the driver and his passenger—that ended with the passenger’s death. Lewis, 523 U.S. at 836–37. Even though the sole question before the Supreme Court was whether the officer had violated the passenger’s rights, the Court applied the purpose-to-harm

ESTATE OF SOAKAI V. ABDELAZIZ 13

test by considering the officer’s intentions and actions only in relation to the driver:

[The officer] was faced with a course of lawless behavior for which the police were not to blame. They had done nothing to cause [the driver’s] high-speed driving in the first place, nothing to excuse his flouting of the commonly understood law enforcement authority to control traffic, and nothing (beyond a refusal to call off the chase) to encourage him to race through traffic at breakneck speed forcing other drivers out of their travel lanes. [The driver’s] outrageous behavior was practically instantaneous, and so was [the officer’s] instinctive response. While prudence would have repressed the reaction, the officer’s instinct was to do his job as a law enforcement officer, not to induce [the driver’s] lawlessness, or to terrorize, cause harm, or kill.

Id. at 855. Indeed, the Supreme Court did not comment on whether the officer might have intended to harm the passenger specifically. Lewis thus established that an officer can—though the officer in Lewis ultimately did not—violate the substantive due process rights of one individual by chasing another for illegitimate purposes. If Lewis left any doubt regarding bystanders, we have since resolved it. In Moreland v. Las Vegas Metropolitan Police Department, 159 F.3d 365 (9th Cir. 1998), officers accidentally shot and killed a bystander while attempting to incapacitate an active shooter. Id. at 368–69. Reviewing the substantive due process claim brought by the bystander’s family, we held that Lewis’s test applies, not just to “high- speed police chases,” but wherever “an officer inadvertently harm[s] a bystander while responding to a situation in which the officer [is] required to act quickly to prevent an individual from threatening the lives of others.” Id. at 372 (emphases added). Defendants’ reading of our caselaw would render that statement nonsensical. How could an officer be said to have intended to harm the person that they “inadvertently” injured? Further contradicting Defendants’ argument, our purpose-to-harm analysis in Moreland did not consider only the officers’ intentions with respect to the bystander. Instead, we affirmed the grant of summary judgment to the officers because they (1) had sought to harm the suspect for legitimate reasons and (2) did not intend to harm the bystander at all. See id. at 373 (noting that the plaintiffs neither “dispute[d] that [the officers] w[ere] entitled to use deadly force to halt the gunfight,” nor contended that the officers “intended to harm [the bystander]”). The decisive question was, in other words, whether the officers had intended to harm someone—rather than the bystander, specifically—for reasons unrelated to a legitimate law enforcement objective. In Onossian, we clarified that Moreland’s reading of Lewis applies equally in car-chase cases. See Onossian, 175 F.3d at 1172 (“It is a small step from applying Lewis to a bystander harmed in a gunfight to applying it to another driver harmed in the very situation in which the Lewis test originated.”). Accordingly, we held that bystanders may recover for injuries caused by a high-speed chase, but only

ESTATE OF SOAKAI V. ABDELAZIZ 15

if they satisfy the purpose-to-harm test. 4 Id. at 1171–72. And when addressing the merits, we followed Lewis’s and Moreland’s leads by looking for “evidence that [the] deputies . . . intended to cause harm to anyone,” id. at 172 (emphasis added), rather than to any one specific individual. Nearly a decade later, we reaffirmed that approach in another car-chase case involving an injured bystander. See Bingue, 512 F.3d at 1177 (asking whether the officer “acted with an intent to harm,” without specifying a particular object of that intent). Adopting Defendants’ position would require us to overrule Onossian and Bingue, which, of course, we cannot do. See, e.g., In re Complaint of Ross Island Sand & Gravel, 226 F.3d 1015, 1018 (9th Cir. 2000) (per curiam) (“A three judge panel of this court cannot overrule a prior decision of this court.”). Because officers generally must follow a route chosen by the fleeing suspect, officers engaged in a high- speed chase presumably cannot predict precisely where the pursuit will take them or, more importantly, who specifically will be endangered when they get there. We thus have difficulty imagining a scenario in which an officer forms the intent to harm a bystander in the brief moments during which

4 It was suggested during oral argument that this decision was merely dictum because the plaintiffs in Onossian did not win. We disagree for two reasons. First, a reasoned decision on an issue actually presented on appeal creates binding law, “regardless of whether [addressing the issue] was in some technical sense ‘necessary’ to our disposition of the case.” Barapind v. Enomoto, 400 F.3d 744, 751 (9th Cir. 2005) (en banc) (per curiam). The decision identified in Onossian was certainly reasoned; indeed, half of the opinion is dedicated to it. Second, the decision was technically necessary to the result. By deciding that Lewis’s purpose-to- harm test applied to claims brought by bystanders, the panel determined the applicable legal standard—a step the panel was required to take before it could address the merits. the bystander is in their line of sight. Given that reality, adopting Defendants’ rule would make it all but impossible for any bystander to recover for injuries caused by a high- speed police chase. We have held, however, that “a bystander injured in a high-speed police chase ‘must show that the behavior of the police in his case meets the Lewis standard.’” Bingue, 512 F.3d at 1175 (alterations adopted) (quoting Onossian, 175 F.3d at 1172). And implicit within that holding is the idea that Lewis’s standard can actually be satisfied in at least some cases involving bystanders. Even were we permitted to set our caselaw aside, Defendants’ position would remain flawed. The fundamental question for the purpose of deciding whether Plaintiffs have stated a substantive due process claim is whether Defendants’ alleged conduct shocks the conscience. Porter, 546 F.3d at 1137. “[C]onduct intended to injure in some way unjustifiable by any government interest” sits at the far “end of the culpability spectrum” and, therefore, “is the sort of official action most likely to rise to the conscience-shocking level.” Lewis, 523 U.S. at 849. We see no reason to think that conduct is any less shocking when it injures someone other than the intended target, particularly when harm to a third party is a clear, known risk and is entirely foreseeable. 5 We therefore hold that Plaintiffs have stated a substantive due process claim by plausibly alleging that they, as bystanders, were injured when Defendants engaged

5 Suppose that an officer walks into a crowd and shoots at an unarmed civilian purely for the purpose of causing pain. Surely, such conduct would shock the conscience whether the officer hits the intended target or instead strikes a bystander standing a few feet away.

ESTATE OF SOAKAI V. ABDELAZIZ 17

in a high-speed chase for the purpose of harming the fleeing suspect without a legitimate law enforcement objective. 2. Clearly Established Law For many of the same reasons discussed above, we also hold that Defendants violated clearly established law. The “unlawfulness of an officer’s conduct [is] ‘clearly established’” if, “at the time of the officer’s conduct, the law was sufficiently clear that every reasonable official would understand that what he is doing is unlawful.” Hampton, 83 F.4th at 769 (quoting District of Columbia v. Wesby, 583 U.S. 48, 63 (2018)) (some internal quotation marks omitted). Lewis, Moreland, and Onossian provided such clarity well before June 25, 2022—the date of the alleged car chase. To recap, Lewis established that conducting a high-speed chase “with [the] intent to harm suspects physically or to worsen their legal plight” violates substantive due process. Lewis, 523 U.S. at 854. Moreland held that Lewis’s standard applies when “an officer inadvertently harm[s] a bystander” in other high-pressure situations. Moreland, 159 F.3d at 372 (emphasis added). And Onossian made clear that Moreland’s interpretation of Lewis governs car-chase cases involving injured bystanders. Onossian, 175 F.3d at 1172. Defendants make much of the fact that neither the Supreme Court nor our court has ruled in favor of a bystander injured in a high-speed chase when addressing the merits of a substantive due process claim. But whether the plaintiffs in prior cases succeeded in establishing a substantive due process claim is beside the point. Instead, we must determine whether the law provided “fair warning that [Defendants’ alleged] conduct [was] unconstitutional.” Ballentine v. Tucker, 28 F.4th 54, 66 (9th Cir. 2022) (quoting Ellins v. City of Sierra Madre, 710 F.3d 1049, 1064 (9th Cir. 2013)) (internal quotation marks omitted). And here, the constitutional rule set forth in Lewis, Moreland, and Onossian “appl[ies] with obvious clarity” to the conduct alleged in this case. Dodge v. Evergreen Sch. Dist. #114, 56 F.4th 767, 784 (9th Cir. 2022); see also A.D., 712 F.3d at 454–55 (explaining that, given controlling precedent, “it would be ‘clear to a reasonable officer’ that killing a person with no legitimate law enforcement purpose violates the Constitution” (quoting Brosseau v. Haugen, 543 U.S. 194, 199 (9th Cir. 2004))). B. State-Created Danger Claim As an alternative, narrower theory of relief, Plaintiffs allege that Defendants’ failure to summon aid after the crash delayed Plaintiffs’ receiving medical treatment, resulting in additional harm to Plaintiffs. The claim is narrow because it addresses only those additional harms that Plaintiffs would not have suffered had Defendants provided or summoned aid right after the crash, not the injuries caused by the crash itself. And the claim is an alternative one because, if Plaintiffs prevail on their purpose-to-harm claim, they will be entitled to recover for all injuries sufficiently connected to the crash—including both their initial injuries and those caused by delayed medical treatment. See Borunda v. Richmond, 885 F.2d 1384, 1389 (9th Cir. 1988) (“A plaintiff who establishes liability for deprivations of constitutional rights actionable under 42 U.S.C. § 1983 is entitled to recover compensatory damages for all injuries suffered as a consequence of those deprivations.”). Because the state-created danger claim is relevant only if Defendants fail to prove the primary theory, described above, we will assume in this part of the opinion that Defendants did not give chase with a purpose to harm

ESTATE OF SOAKAI V. ABDELAZIZ 19

unrelated to a legitimate law enforcement objective. With that in mind, we hold that Defendants are not entitled to qualified immunity with respect to Plaintiffs’ state-created danger claim. 1. Constitutional Violation The Fourteenth Amendment “generally does not confer any affirmative right to governmental aid.” Patel v. Kent Sch. Dist., 648 F.3d 965, 971 (9th Cir. 2011). But there are exceptions to that rule, including, as relevant here, the “state- created danger” exception, which applies when the state fails “to protect a plaintiff that it affirmatively place[d] in danger by acting with deliberate indifference to a known or obvious danger.” Sinclair v. City of Seattle, 61 F.4th 674, 680 (9th Cir. 2023) (quoting Martinez v. City of Clovis, 943 F.3d 1260, 1271 (9th Cir. 2019)) (internal quotation marks omitted). To make use of the state-created danger exception, a plaintiff must satisfy two requirements, both of which relate to the defendant-officer’s conduct. First, the plaintiff must establish that the officer’s affirmative conduct exposed the plaintiff “to a foreseeable danger that she would not otherwise have faced.” Martinez v. High, 91 F.4th 1022, 1028 (9th Cir.), cert. denied, No. 24-130, 145 S. Ct. 547 (2024). Second, the plaintiff must show that the officer acted with “deliberate indifference to a known or obvious danger.” Id. (quoting Murguia v. Langdon, 61 F.4th 1096, 1111 (9th Cir. 2023)) (internal quotation marks omitted). Those two requirements can fit together in different ways. Often, the affirmative conduct at issue under the exception’s first prong will be the same conduct that illustrates the officer’s deliberate indifference for purposes of the second. In such circumstances, officers are held liable for “demonstrat[ing] deliberate indifference in creating a danger.” Bracken v. Okura, 869 F.3d 771, 779 n.7 (9th Cir. 2017); see also Kennedy v. City of Ridgefield, 439 F.3d 1055, 1065 (9th Cir. 2006) (concluding that sufficient evidence existed to suggest that the officer “acted deliberately and indifferently to the danger he was creating”). By contrast, a substantive due process violation also can arise if (1) an officer’s affirmative conduct exposes the plaintiff to danger, and (2) the officer then acts with deliberate indifference when “responding”—or, as the case may be, failing to respond—to that danger. Bracken, 869 F.3d at 779 n.7. Simply put, “[w]hen an officer’s affirmative conduct creates a foreseeable risk of harm to the plaintiff, the officer will be liable for failing to intercede if the officer demonstrates ‘deliberate indifference’ to the plaintiff’s plight.” Id. at 778–79; see also L.W. v. Grubbs, 92 F.3d 894, 896 (9th Cir. 1996) (“[D]eliberate indifference on the part of the responsible official, to the safety of employees in the presence of known danger, created by official conduct, is sufficient to establish a due process violation . . . .” (emphasis added)). Bracken exemplifies the second category of cases. There, an officer prevented the plaintiff from leaving an event. Bracken, 869 F.3d at 775. Private security guards also working at the event then arrived and began to assault the plaintiff. Id. The officer still refused to let the plaintiff leave, and the security guards eventually tackled the plaintiff to the floor. Id. at 779. Reviewing the plaintiff’s substantive due process claim, we concluded that the doctrine’s first prong was satisfied because the officer’s “active participat[ion] in the incident” exposed the plaintiff to danger that he would not have faced had the officer allowed

ESTATE OF SOAKAI V. ABDELAZIZ 21

him to leave. Id. (emphasis added). We then went on to hold that a reasonable jury could find that the officer acted with deliberate indifference when he “chose to do nothing” as the security guards continued their assault. Id. at 780 (emphasis added). In other words, the officer created the danger by preventing the plaintiff’s departure, and the officer then was deliberately indifferent by failing to intervene in the assault. Wood v. Ostrander, 879 F.2d 583 (9th Cir. 1989)— where we first considered the doctrine of state-created danger—illustrates the same principle. The officer in that case pulled a car over, had it towed, and arrested its intoxicated driver. Id. at 586. The officer then drove away, stranding the driver’s passenger in an area with a high crime rate, and an unknown man later attacked the passenger. Id. As in Bracken, the officer’s affirmative conduct—arresting the driver and impounding the car—exposed the passenger to risks that she would not have faced otherwise. And by driving off on his own, without helping the passenger in any way, the officer failed to address that risk. We thus held that, by showing that the officer did nothing to ameliorate the danger he had created, the passenger had “raised a genuine issue of fact tending to show that [the officer] acted with deliberate indifference to [the passenger’s] interest in personal security under the [F]ourteenth [A]mendment.” Id. at 588. Plaintiffs’ state-created danger theory fits the mold of Bracken and Wood. Plaintiffs argue that they have satisfied both of the doctrine’s requirements by plausibly alleging that Defendants affirmatively created danger by initiating a car chase that led to a crash and then acting with deliberate indifference to Plaintiffs’ worsening medical condition by failing to summon help. Accepting Plaintiffs’ allegations as true and drawing all reasonable inferences in Plaintiffs’ favor, we agree. a. Affirmative Conduct To satisfy the state-created danger exception’s first prong, Plaintiffs must plausibly allege that Defendants’ affirmative actions (1) placed Plaintiffs in a worse position than they would have occupied had Defendants not acted at all; (2) created or exposed Plaintiffs to an actual and particularized danger; and (3) resulted in foreseeable harm to Plaintiffs. Polanco v. Diaz, 76 F.4th 918, 926 (9th Cir. 2023), cert. denied, No. 23-842, 144 S. Ct. 2519 (2024). The first of those elements requires little discussion. Plaintiffs allege that Defendants “sped after the suspect” without alerting the suspect to pull over by turning on their lights and sirens. We can plausibly infer from those allegations that, in the absence of Defendants’ affirmative actions, the suspect would not have raced through the city and lost control of his vehicle and, therefore, that Plaintiffs would not have required urgent medical attention to keep their conditions from deteriorating further. So, even if Defendants initiated the chase for a legitimate purpose, Defendants undoubtedly “increased the level of danger” faced by Plaintiffs “above the counterfactual baseline level of danger that would have existed without [Defendants’] intervention.” Sinclair, 61 F.4th at 682. 6

6 Not all car crashes that follow a pursuit will satisfy this element. For example, if a fleeing suspect was already driving erratically and if the officers’ conduct did not increase the risk of a crash, then this element likely would not be met. No liability would attach if Defendants’ actions “did not place [Plaintiffs] in any worse position than they would have been in had [Defendants] not [done anything] whatsoever.” Johnson v. City of Seattle, 474 F.3d 634, 641 (9th Cir. 2007).

ESTATE OF SOAKAI V. ABDELAZIZ 23

Foreseeability of harm is also easily addressed. It is entirely predictable that allowing seriously wounded individuals to go without aid for longer than necessary would increase the risk of further injury or death. Cf. Maxwell v. County of San Diego, 708 F.3d 1075, 1083 (9th Cir. 2013) (“It was obvious that delaying a bleeding gun shot victim’s ambulance increased the risk of death.”). Regarding the last remaining element, Defendants argue that they did not create a “particularized” danger because the high-speed chase “threatened the safety of the public at large.” Were we focused on the danger wrought by the car chase alone, Defendants would have a valid point. See Polanco, 76 F.4th at 927 (“Affirmative state action that exposes a broad swath of the public to ‘generalized dangers’ cannot support a state-created-danger claim.”). As we have explained, however, Plaintiffs’ state-created danger claim targets not the injuries caused by the crash itself but the additional risk faced by Plaintiffs after the crash due to delayed medical treatment. Only a small and distinct group—those few people injured by the collision—had to contend with that danger. Plaintiffs therefore rely on a sufficiently particularized danger. See id. (explaining that a danger can be particularized if experienced by “a ‘discrete and identifiable group’” (quoting Sinclair, 61 F.4th at 683)). b. Deliberate Indifference Next, we ask whether Defendants showed deliberate indifference in the presence of the known danger that they created. Bracken, 869 F.3d at 779. In this context, we employ a subjective “deliberate indifference” standard. Polanco, 76 F.4th at 928. To satisfy that standard, Plaintiffs must show that Defendants knew that their intentional actions would expose Plaintiffs to an unreasonable risk. Murguia, 61 F.4th at 1117 n.16. Put differently, Plaintiffs must allege facts from which we can plausibly infer that Defendants “kn[ew] that something was going to happen, but ‘ignored the risk and exposed [Plaintiffs] to it anyway.’” City of Clovis, 943 F.3d at 1274 (quoting Hernandez v. City of San Jose, 897 F.3d 1125, 1135 (9th Cir. 2018)). Plaintiffs have carried their burden. Defendants allegedly saw the “carnage, injuries[,] and death” caused by the collision but did not render aid or call for help, even though the police department had a policy—and had trained Defendants to follow the policy—that officers “provide and summon emergency medical services for injured persons.” Instead, Defendants allegedly kept driving, without turning on their sirens or lights, before doubling back and pretending to arrive at the scene for the first time after hearing other first responders arrive. From those allegations, we can reasonably infer that Defendants saw that Plaintiffs needed immediate medical attention, knew from their training—and common sense—that the danger to Plaintiffs would increase the longer they went without help, and—like the officer in Bracken—still “chose to do nothing about it.” Bracken, 869 F.3d at 780. Defendants characterize Plaintiffs’ reliance on the state- created danger doctrine as an attempt to circumvent the rule that only a purpose to harm, and not deliberate indifference, can support a substantive due process violation in cases involving high-speed chases. Some of the broad phrasing in our prior decisions appears to support that assertion. See, e.g., Bingue, 512 F.3d at 1170–71 (holding that “police officers involved in all high-speed chases are entitled to qualified immunity . . . unless the plaintiff can prove that the officer acted with a deliberate intent to harm”). But those cases focused on injuries that resulted from an officer’s

ESTATE OF SOAKAI V. ABDELAZIZ 25

“decision to join the chase in the first place or the serial decisions about how best to pursue the suspect.” Id. at 1176. Requiring a more demanding showing of fault before penalizing such choices makes sense because all of those decisions must be made without “the luxury of delay.” Id. The same logic does not apply, however, to Defendants’ decision to leave the scene of the crash, a choice Defendants made only after the chase was over and the suspect was no longer mobile. On the contrary, we can readily infer from Plaintiffs’ allegations that Defendants had time to deliberate before driving away. Relatedly, the unique aspects of this case ensure that our decision will not undermine the holdings of Onossian and Bingue. Deliberate indifference remains a “stringent standard of fault,” Murguia, 61 F.4th at 1111 (quoting Patel, 648 F.3d at 974), requiring a “culpable mental state,” id. Not even gross negligence will do. Id. Failing to aid victims of a car crash will thus rarely constitute deliberate indifference; if, for instance, an officer did not summon emergency services because the officer was distracted by the need to locate the suspect, did not see the victims, or was injured, the officer almost certainly would lack the “culpable mental state” required. Id. In the universe of high-speed chase cases, this one, with its particularly unusual allegations, is an outlier. We therefore hold that Plaintiffs have stated a claim under the state-created danger exception. 2. Clearly Established Law We also hold that, if Plaintiffs’ allegations are true, Defendants violated clearly established law by acting with deliberate indifference to the injuries that resulted from the collision that Defendants affirmatively helped to cause. As Defendants point out, Plaintiffs do not identify—nor are we aware of—a controlling case applying the state- created danger theory to injuries suffered by bystanders after a high-speed police chase. But “[t]here need not be a case directly on point for a right to be clearly established.” City of Clovis, 943 F.3d at 1275; see also Hope v. Pelzer, 536 U.S. 730, 741 (2002) (“[O]fficials can still be on notice that their conduct violates established law even in novel factual circumstances.”). We ask only whether existing precedent has placed the constitutional question “beyond debate.” City of Clovis, 943 F.3d at 1275 (quoting Shafer v. County of Santa Barbara, 868 F.3d 1110, 1117 (9th Cir. 2017)). Here, two lines of cases settle the debate. First, in cases like Wood and Bracken, we established that officers violate substantive due process when they affirmatively place an individual in danger and then, with deliberate indifference, do nothing to address that danger. See Wood, 879 F.2d at 588 (allowing the state-created danger claim to proceed when the officer deprived the passenger of her ride and then “left [her] by the side of the road at night in a high-crime area”); Bracken, 869 F.3d at 778–79 (“When an officer’s affirmative conduct creates a foreseeable risk of harm to the plaintiff, the officer will be liable for failing to intercede if the officer demonstrates ‘deliberate indifference’ to the plaintiff’s plight.”). Also relevant are cases like Maxwell, in which we held that “[i]mpeding access to medical care” both (1) “amounts to leaving a victim in a more dangerous situation” and (2) constitutes deliberately indifferent conduct when it is “obvious” that delaying a seriously wounded individual’s access to treatment will “increase[] the risk of death.” Maxwell, 708 F.3d at 1082–83; cf. Penilla v. City of Huntington Park, 115 F.3d 707, 710 (9th Cir. 1997) (per

ESTATE OF SOAKAI V. ABDELAZIZ 27

curiam) (affirming denial of qualified immunity when officers allegedly took affirmative actions that “made it impossible for anyone to provide emergency medical care to [the decedent]”). Considered together, the foregoing authorities show that Defendants’ alleged conduct violated a clearly established right. See Polanco, 76 F.4th at 930 n.8 (“We routinely rely on the intersection of multiple cases when holding that a constitutional right has been clearly established.”). True, Maxwell and Penilla involved affirmative steps taken to delay aid that was already underway. In Maxwell, for example, the officers “prevent[ed] [the injured individual’s] ambulance from leaving.” Maxwell, 708 F.3d at 1082. But those cases nevertheless put officers on notice that hampering an individual’s access to emergency treatment can constitute a substantive due process violation when the “affirmative conduct” prong of the state-created danger exception is also satisfied. And Bracken and Wood make clear that an officer’s inaction—such as failing to provide aid—can violate the Due Process Clause if the officer previously put another in danger, even if the officer’s earlier danger-creating conduct served legitimate law enforcement objectives. See Bracken, 869 F.3d at 775 (reasoning that an officer affirmatively placed the plaintiff in danger when the officer acted with a legitimate law enforcement purpose); Wood, 879 F.2d at 588 (holding that the officer was not entitled to summary judgment when he may have endangered the plaintiff after carrying out legitimate law enforcement purposes).

RESPONSE TO THE DISSENTING OPINION

The dissenting opinion’s heated rhetoric ignores what this opinion says and what our precedents provide. We answer briefly the dissenting opinion’s two central unsupported claims. First, the dissenting opinion reads our precedent to require that the intent to harm must be the intent to harm the injured bystander specifically. Dissent at 34–40. Not so. In Moreland, we affirmed the summary judgment entered in the officers’ favor “because the officers were responding to the extreme emergency of public gunfire and did not intend to commit any harm unrelated to the legitimate use of force necessary to protect the public and themselves.” 159 F.3d at 373 (emphasis added). In other words, the Moreland plaintiffs lost because there was no evidence that the officers acted with an impermissible purpose to harm, as to either the suspect (whom the officers intended to shoot) or the bystander (whom the officers shot accidentally). And in Onossian, a car-chase case in which we applied Moreland’s interpretation of Lewis, we asked simply whether the deputies “intended to cause harm to anyone.” Onossian, 175 F.3d at 1172 (emphasis added). Those cases, on their own, clearly establish the law regarding bystanders. But we also note that in Porter, 546 F.3d at 1140, we quoted with approval Judge McKee’s concurring opinion in Davis v. Township of Hillside, 190 F.3d 167 (3rd Cir. 1999)—a case brought by a bystander who was injured as a result of a high- speed police car chase—to explain that “[i]t is the intent to inflict force beyond that which is required by a legitimate law enforcement objective that shocks the conscience and gives rise to liability under § 1983.” In that opinion, Judge McKee explained that “if the record supported a finding that police gratuitously rammed [the fleeing suspect’s] car, and if [the] plaintiff[-bystander] properly alleged that they did so to injure or terrorize [the suspect], liability could still attach

ESTATE OF SOAKAI V. ABDELAZIZ 29

under Lewis.” Id. at 172–73 (McKee, J., concurring) (emphasis added). Second, the dissenting opinion alleges that we have ruled that officers are required to render or summon medical aid for civilians who are harmed by private actors. Dissent at 43–45. Again, not so. Setting aside the dissenting opinion’s more general disagreement with our precedents concerning the state-created danger doctrine, we have faithfully followed those precedents. Here, according to the complaint, after intentionally causing the crash, the officers saw the resulting “carnage, injuries[,] and death” but decided to drive away and to return later, pretending to arrive at the scene for the first time. That sequence of events is no ordinary failure to render aid. Instead, the complaint alleges that Defendants deliberately abandoned Plaintiffs in a dangerous situation which, because of the affirmative actions taken by the Defendants in the lead-up to the crash, was of Defendants’ own making. See Bracken, 869 F.3d at 779–80 (rejecting the assertion of qualified immunity when a jury could find that (1) the defendant “engaged in affirmative conduct that exposed [the plaintiff] to foreseeable harm”; and (2) the defendant knew that plaintiff was being harmed “and deliberately chose to do nothing about it”).

CONCLUSION

We hold that Defendants are not entitled to qualified immunity with respect to either of Plaintiffs’ theories of substantive due process liability. Given the unique facts of this case, we also emphasize that “our decision at the motion-to-dismiss stage sheds little light on whether the government actors might ultimately be entitled to qualified immunity” at later stages of the proceeding. Keates v. Koile, 883 F.3d 1228, 1235 (9th Cir. 2018). We thus express no opinion on what the district court might conclude at summary judgment or, should the case proceed to trial, what a jury might find.

AFFIRMED.

BUMATAY, Circuit Judge, dissenting:

No doubt this case presents jarring facts. According to the complaint, two Oakland Police Department officers, Walid Abdelaziz and Jimmy Marin-Coronel, began chasing a suspect from an illegal car rally. The officers followed the suspect’s car through the highly populated streets of Oakland—at speeds up to 100 mph. Even more, the officers engaged in a “ghost chase”—they pursued the suspect without their lights and sirens, and they didn’t radio in the chase to police dispatch. They did this because the Oakland Police Department prohibits police chases with rare exception. Sadly, the suspect lost control of his vehicle and crashed into cars and motorcycles parked by a late-night taco truck. Several bystanders waiting for food at the taco truck were struck; Lolomania Soakai died from his injuries. The officers witnessed the crash but did not stop or summon medical aid. The officers came on scene—acting as if they had only just arrived—after they heard other officers had already responded. The officers were overheard commenting that they hoped the suspect had died in the crash. The family of Soakai and other injured bystanders sued (“Plaintiffs”). As shocking as these allegations appear, we must always adhere to our constitutional role. That means following

ESTATE OF SOAKAI V. ABDELAZIZ 31

established law and not grasping at rulings to reach certain outcomes. As I’ve said previously, “[f]aced with tragic facts, . . . we may be tempted to expand the scope of constitutional rights . . . . But our job is to look to the text and history of the Constitution for the scope of constitutional remedies—not simply to make good the wrong done.” Murguia v. Langdon, 73 F.4th 1103, 1103 (9th Cir. 2023) (Bumatay, J., dissenting from denial of rehearing en banc) (simplified). Indeed, qualified immunity’s “clearly established” standard “protects the balance between vindication of constitutional rights and government officials’ effective performance of their duties” by making sure that government officials can “reasonably anticipate when their conduct may give rise to liability for damages.” Reichle v. Howards, 566 U.S. 658, 664 (2012) (simplified). “A clearly established right is one that is sufficiently clear that every reasonable official would have understood that what he is doing violates that right.” Mullenix v. Luna, 577 U.S. 7, 11 (2015) (per curiam) (simplified). “This exacting standard gives government officials breathing room to make reasonable but mistaken judgments by protecting all but the plainly incompetent or those who knowingly violate the law.” City & Cnty. of San Francisco v. Sheehan, 575 U.S. 600, 611 (2015) (simplified). Thus, qualified immunity weeds out claims premised on novel or opaque theories of constitutional violation. Yet the majority greenlights exactly that sort of claim—offering a string of unprecedented rulings untethered from the Constitution. First, the majority adopts a brand-new theory of substantive due process—contrary to precedent and to the Supreme Court’s admonition against such judicial overreach. See Dobbs v. Jackson Women’s Health Org., 597 U.S. 215, 239–40 (2022). For the first time, the majority rules that a bystander may assert a substantive due process claim against an officer if the bystander can show that the officer intended to harm someone else. In other words, the majority contends that the bystander Plaintiffs here showed a clearly established due process violation even though the allegations establish that the police officers intended to harm only the suspect—not the bystanders. But neither the Supreme Court nor the Ninth Circuit has ever endorsed this theory. So it’s no wonder that the majority can’t point to a single Supreme Court or Ninth Circuit opinion stating that intent to harm someone else violates the Due Process Clause. In fact, the Ninth Circuit precedent goes completely the other way. More than 25 years ago, we concluded that similar bystander plaintiffs “failed to state a viable substantive due process claim because” they failed to satisfy “the controlling question of whether [the officer] acted with a purpose to harm Douglas [the bystander] that was unrelated to his attempt to stop the [suspect in the case] from endangering others.” Moreland v. Las Vegas Metro. Police Dept., 159 F.3d 365, 373 (9th Cir. 1998) (simplified). So we definitively held that bystanders must establish intent to harm the bystanders to sustain a due process claim. To get around this, the majority completely misreads Moreland and elevates opaque legalese, convoluted innuendo, a single- judge concurrence, and snippets from different cases to “clearly established” law. Simply unprecedented. Second, the majority expands the state-created-danger doctrine to create a new constitutional duty requiring law enforcement officers to render or summon medical aid for civilians harmed by private actors under certain circumstances. Under this judicially created doctrine, government officials violate substantive due process if they

ESTATE OF SOAKAI V. ABDELAZIZ 33

affirmatively place plaintiffs in danger. The Supreme Court has never recognized the state-created-danger doctrine, and its roots are both ahistorical and atextual. So we shouldn’t casually expand the doctrine. That means following our precedent closely. And at a minimum, what’s needed to make a claim under the doctrine is “affirmative conduct” by a state actor that exposed plaintiffs to “actual, particularized danger.” Polanco v. Diaz, 76 F.4th 918, 926 (9th Cir. 2023) (simplified). But under the majority’s novel theory of state- created danger, no state action with a particularized danger is necessary. All that’s needed now is state action without particularized danger coupled with state inaction with apparent particularized danger. So the majority creates a state duty to render aid whenever a private actor harms civilians if police officers acted in any way in the causal chain of harm. This is a confusing expansion of a dubious doctrine. Only by reaching novel holding after novel holding can the high standard of “clearly established law” be overcome. This was not our role. Rather than hack together unprecedented rulings to create not one—but two—new due process rights, we should have simply followed the law and precedent and granted qualified immunity. I respectfully dissent. I. Due Process Claim From the Police Chase To begin, the Supreme Court has endorsed only a narrow substantive due process claim for suspects injured during a police chase. In County of Sacramento v. Lewis, 523 U.S. 833 (1998), the Court held that a police officer with “a purpose to cause harm unrelated to the legitimate object of arrest” violates due process by causing injuries to a suspect during a high-speed chase. Id. at 836. Thus, it is clearly established that “a police officer who acted with the purpose to harm a civilian, unrelated to the legitimate law enforcement objectives of arrest, self-defense, or the defense of others, violated the Fourteenth Amendment due process clause.” A.D. v. Cal. Highway Patrol, 712 F.3d 446, 454 (9th Cir. 2013). The majority denies qualified immunity to the Officers on Plaintiffs’ due process claim based on the injuries they sustained after the Officers pursued the suspect during the “ghost chase.” In other words, the majority holds that a bystander may assert a due process violation if a police officer intends to harm someone else as long as the bystander is injured somewhere in the causal chain. But that’s wrong for two reasons. First, even assuming that Plaintiffs adequately alleged an intent to harm the suspect unrelated to a legitimate objective, there’s no basis under the Constitution to find liability for injuries to bystanders from the officers’ conduct. Second, it wasn’t clearly established law at the time of the accident that intent to harm a suspect is enough to press a due process claim for injuries to bystanders. First, the Constitution doesn’t support expanding due process protections to a bystander harmed by an officer who intends to harm someone else. The majority reasons that high-speed car chases endanger “the fleeing driver and . . . the officers,” and “passengers and bystanders.” Maj. Op. 12. It then claims it is “illogical” to distinguish between the two “dangers” in deciding the scope of the Due Process Clause. Id. That’s because, to the majority, harm to bystanders is a “clear, known risk and is entirely foreseeable.” Id. at 16.

ESTATE OF SOAKAI V. ABDELAZIZ 35

The majority ignores that this is constitutional law—not tort law. The language of “risks” and “foreseeability” is the language of negligence—not the Due Process Clause. Of course, a police chase causing injuries to a bystander may give rise to a tort action against the city. See, e.g., City of Sacramento v. Superior Ct., 131 Cal. App. 3d 395 (Ct. App. 1982); Est. of Aten v. City of Tucson, 817 P.2d 951 (Ariz. Ct. App. 1991). But the question here is whether the Constitution protects against such unintended injuries. And that answer is clearly no. As Lewis itself recognized, we can’t “demote[]” the Constitution to a “font of tort law.” 523 U.S. at 847 n.8. Lewis understood that we “need to preserve the constitutional proportions of constitutional claims[.]” Id. Lewis thus stressed that the Due Process Clause can’t be used “to supplant traditional tort law” and it may not “lay[] down [the] rules of conduct to regulate liability for injuries that attend living together in society.” Id. at 848 (simplified). Thus, “the Constitution does not guarantee due care on the part of state officials” and “liability for negligently inflicted harm is categorically beneath the threshold of constitutional due process.” Id. It is only “deliberate decisions of government officials to deprive a person of life, liberty, or property” that offends the Due Process Clause. Id. (simplified). Because an officer doesn’t deliberately intend for any harm to the “life, liberty, or property” of bystanders if the officer’s intent is trained on a suspect, at most, the officer shows a reckless disregard for bystanders’ lives by engaging in a police chase. And Lewis expressly rejected the “reckless disregard for life” standard when it comes to due process claims from a high-speed chase. Id. at 854. Lewis then confirmed that, for a due process violation to occur, an officer must have an intent to “terrorize, cause harm, or kill” the person. Id. at 855. That doesn’t happen when an officer’s actions inadvertently injure a bystander. So the majority was wrong to endorse this newfound facet of substantive due process. Indeed, the Ninth Circuit has already answered this question and concluded that, to state a due process claim, the “controlling question” is whether the officer “acted with a purpose to harm” the bystander “that was unrelated to” a legitimate law enforcement purpose. See Moreland, 159 F.3d at 373. Second, given how this novel theory of due process conflicts with Supreme Court and Ninth Circuit precedent, there’s no way it is “clearly established law.” Recall the “clearly established law” standard “protect[s] all but the plainly incompetent or those who knowingly violate the law.” Sheehan, 575 U.S. at 611 (simplified). But under the majority’s reading of the standard, it protects no one except those with four law clerks and a Westlaw subscription. It requires police officers to ignore directly controlling authority, to squint at our caselaw, and to string together creative interpretations across multiple cases. Indeed, look at the majority’s tortured path to get to clearly established law:

To recap, Lewis established that conducting a high-speed chase “with [the] intent to harm suspects physically or to worsen their legal plight” violates substantive due process. Lewis, 523 U.S. at 854. Moreland held that Lewis’s standard applies when “an officer inadvertently harm[s] a bystander” in other high-pressure situations. Moreland, 159 F.3d at 372 (emphasis added). And Onossian made clear that Moreland’s interpretation of Lewis

ESTATE OF SOAKAI V. ABDELAZIZ 37

governs car-chase cases involving injured bystanders. Onossian, 175 F.3d at 1172.

Maj. Op. 17. That’s about as clear as a cloudy day. One would expect that, for something to meet the high standard of clearly established law, the majority could point to a single Supreme Court or Ninth Circuit statement that makes “every reasonable official” understand “that what he is doing violates” the law. Mullenix, 577 U.S. at 11. We have nothing of the sort here. Instead, we have a string cite of opaque legalese. And even then, the majority’s reading of these cases is dubious. As the majority admits, Lewis did not directly approve of a bystander’s due process claim based on an officer’s intent to harm someone else. See Maj. Op. 13 (“[T]he Supreme Court did not comment on whether the officer might have intended to harm the passenger specifically.”). That’s because the issue of bystander liability never came up in that case. Indeed, it’s even a stretch to say that a bystander was part of Lewis. In that case, the police officers were pursuing two teenage boys on a motorcycle—with the passenger being killed, 523 U.S. at 836–37—so it’s not a case involving a completely innocent bystander inadvertently injured by a high-speed chase. See also Onossian v. Block, 175 F.3d 1169, 1171 (9th Cir. 1999) (observing that Lewis itself referred to the two teens as “suspects”). And, as stated earlier, Moreland adopts the opposite conclusion—it endorses that a plaintiff bystander must show intent to harm the bystander before due process liability can attach. Thus, the majority completely misreads the decision to claim it as clearly established law going the other way. Moreland involved a midnight gunfight in a parking lot where officers fired on an armed suspect but shot a bystander, Damon Douglas. 159 F.3d at 367. We first decided what due process standard to apply:

The question we face today is whether [Lewis’s] newly minted explanation of the “shocks the conscience” standard also controls in cases where it is alleged that an officer inadvertently harmed a bystander while responding to a situation in which the officer was required to act quickly to prevent an individual from threatening the lives of others. We conclude that it does.

Id. at 372. From this question, which mostly just describes the facts of the case, the majority sees “bystander liability.” But all this statement does is adopt the legal framework for analyzing the bystander’s due process claim. It says nothing about whether the intent to harm someone else is sufficient for a bystander to bring a due process claim against officers. In answering that question, however, Moreland goes the other way. When applying the legal standard to the facts of the case, Moreland dismissed the case because “Appellants do not contend Burns [the officer] intended to harm Douglas [the bystander], physically or otherwise.” Id. at 373. If there was any doubt, Moreland made it crystal clear: “Appellants have failed to state a viable substantive due process claim because these matters are not material to the controlling question of whether Burns acted with a purpose to harm Douglas that was unrelated to his attempt to stop the male in the parking lot from endangering others.” Id. So the “controlling question” in Moreland was whether the bystander showed that the officer “acted with a purpose to harm” the bystander. Id. That settles it clearly. The majority

ESTATE OF SOAKAI V. ABDELAZIZ 39

thus overturns Moreland in reaching its decision, which it can’t properly do. Next, the majority relies on Onossian. In that case, a family sued sheriff’s deputies after a suspect being chased by deputies crashed into their vehicle. 175 F.3d at 1170–71. Onossian expressly limited itself to two questions—neither of which is our question today. Id. at 1171 (“We must decide two questions in this case.”). The first question was: “[D]oes the Lewis test apply not only to harm caused to those pursued in a high speed chase, but also to harm caused to other drivers?” Id. We answered that question affirmatively— yes, Lewis governs police chases leading to injury of bystanders. And in examining that question, we observed that “the duty of the pursuing police officer is defined generally [in Lewis], without specific reference to the suspect being pursued.” Id. We then read Lewis to provide the following rule: “[I]f a police officer is justified in giving chase, that justification insulates the officer from constitutional attack, irrespective of who might be harmed or killed as a consequence of the chase.” Id. This doesn’t resolve whether bystanders must show intent to harm the bystanders to assert a due process claim. The second question Onossian confronted was: “[D]oes the conduct in this case ‘shock the conscience’ within the meaning of Lewis?” Id. at 1172. We answered no. We held that “no reasonable trier of fact could find that defendants’ actions shock the conscience” because there “is no evidence that deputies . . . intended to cause harm to anyone.” Id. at 1172. The majority latches onto these three words from Onossian—“harm to anyone”—to assert that the Ninth Circuit has definitively ruled for bystander liability. Those words do no such thing. However those three words can be interpreted, they neither serve as a holding or even binding well-reasoned dicta on the issue. Nor are they unambiguous enough to provide clearly established law. Whether intent to harm the suspect satisfies bystander liability was not squarely raised in Onossian. At most, the language meant to convey that the plaintiffs’ allegations didn’t come close to asserting a due process foul because the deputies had no intent to harm anyone, which “insulates the officer[s] from constitutional attack.” Id. at 1171. Finally, the majority belatedly tries to justify its holding based on Porter’s citation to a concurrence by a single out- of-circuit judge. Maj. Op. 28–29. But Porter cited Judge McKee’s concurrence for a simple proposition—that “the intent to inflict force beyond that which is required by a legitimate law enforcement objective . . . ‘shocks the conscience.’” See Porter, 546 F.3d at 1140 (quoting Davis v. Twp. of Hillside, 190 F.3d 167, 172 (3d Cir. 1999) (McKee, J., concurring)). Nowhere did Porter adopt the other parts of Judge McKee’s concurrence on which the majority now relies. While creative, this doesn’t prove clearly established law. We should have granted qualified immunity to the Officers on this new substantive due process claim. Even though the prospect of bystander liability for intent to harm a suspect is unprecedented in its own right, what makes this more extraordinary is the assertion that this novel doctrine is clearly established law—apparently hiding in plain sight among our caselaw. That’s not how we determine clearly established law.

ESTATE OF SOAKAI V. ABDELAZIZ 41

II.

Due Process Claim From Failure to Render Aid Plaintiffs also bring another substantive due process claim revolving around the allegations that the Officers failed to render or summon medical help after they witnessed the suspect crash his car into the vehicles next to the taco truck. Under this theory of due process, the Officers aren’t liable for causing the crash but are for the injuries Plaintiffs suffered from their failure to immediately provide medical attention after the crash. This species of judge-made law is often called the “state-created danger” doctrine—but Plaintiffs argue for its expansion beyond anything courts have recognized. The majority thus wrongly endorses this newfangled claim. The Due Process Clause imposes no “affirmative obligation on the State” to protect a person’s life, liberty, or property, and it serves only as a “limitation on the State’s power to act” rather than a “guarantee of certain minimal levels of safety and security.” DeShaney v. Winnebago Cnty. Dep’t of Soc. Servs., 489 U.S. 189, 195 (1989). So ordinarily Plaintiffs’ claim would fail from the start because they seek to charge government actors with injuries caused by the unnamed suspect. But the state-created-danger doctrine creates “an exception to the rule that the Due Process Clause does not obligate the State to protect its citizens from harm caused by private actors.” Murguia, 73 F.4th at 1103 (Bumatay, J., dissenting from denial of rehearing en banc). Instead, under the doctrine, government actors “may be liable for their roles in creating or exposing individuals to danger they otherwise would not have faced.” Polanco, 76 F.4th at 926 (simplified). If this sounds untethered from the Due Process Clause, it is. The “state-created danger exception finds no support in the text of the Constitution, the historical understanding of the ‘due process of law,’ or even Supreme Court precedent.” Murguia, 73 F.4th at 1104 (Bumatay, J., dissenting from denial of rehearing en banc). It was manufactured out of whole cloth by courts and aggressively expanded by the Ninth Circuit. Given its dubious pedigree, we should reject the doctrine’s “undue expansion” and “align it with the text of the Due Process Clause and Supreme Court precedent to the extent possible.” Id. Rather than rein in the doctrine, the majority continues its ever-expansion—this time creating a new state duty to provide immediate medical aid. To assert a due process claim under the state-created-danger doctrine, plaintiffs must first “allege affirmative conduct on the part of the state” that “exposed” them to “an actual, particularized danger” that they “would not otherwise have faced,” and, second, that the state official “acted with deliberate indifference to that known or obvious danger.” Polanco, 76 F.4th at 926. So the first hurdle for plaintiffs is alleging “affirmative conduct” that exposed them to a “particularized danger.” Id. That’s a tough one for Plaintiffs given that their allegations stem from a car crash—the dangers of which are of the most general kind. Car accidents are an unfortunate but ubiquitous risk on our roadways. They can strike almost anyone at almost any time. It would be nearly impossible to predict where a suspect in flight will crash or who might be injured. And an “[a]ffirmative state action that exposes a broad swath of the public to ‘generalized dangers’ cannot support a state-created-danger claim.” Id. at 927; see also Sinclair v. City of Seattle, 61 F.4th 674, 682 (9th Cir. 2023) (observing that “[a] particularized danger, naturally, contrasts with a general one” and must be “directed at a

ESTATE OF SOAKAI V. ABDELAZIZ 43

specific victim”) (simplified). So even if the Officers played a role in causing the crash, Plaintiffs can’t plausibly allege that the danger was “particularized” to them. Indeed, under the allegations of the complaints, Plaintiffs can’t show that Officers knew of the particular danger to them because the Officers didn’t even know of their existence until after the crash. Even the majority admits that Plaintiffs can’t assert a state-created-danger claim based on the car crash. See Maj. Op. 23 (“Were we focused on the danger wrought by the car chase alone, Defendants would have a valid point.”). But the majority waves away the flaws in Plaintiffs’ claim by instituting a novel state affirmative duty to render or summon medical aid. Id. (“Plaintiffs’ state-created danger claim targets not the injuries caused by the crash itself but the additional risk faced by Plaintiffs after the crash due to delayed medical treatment.”). In the majority’s view, then, the Officers are liable under the state-created-danger doctrine because they engaged in “affirmative conduct” after the crash—by doing nothing. But inaction is not affirmative conduct by definition. See Merriam-Webster’s Dictionary of Law 17 (1996) (defining “affirmative” as “resulting from an intentional act” or “involving or requiring application of effort”). Merely failing to help or summon help—however heartless—doesn’t amount to affirmative governmental conduct sufficient to sustain a due process claim. And the majority’s new reading of due process flatly contradicts our precedent. Under our precedent, we ask whether state actors “placed the plaintiff in a worse position than he would have been in had the state not acted at all.” Polanco, 76 F.4th at 926 (simplified). But when officers fail to render medical assistance, they are, by definition, “not act[ing] at all.” So the majority rewrites the inquiry. Now, we must ask—could officers have helped plaintiffs if they had acted? The majority thus transforms the Due Process Clause into a mandate of government assistance against harm from private actors. The majority’s conception of the Due Process Clause defies both its text and historical understanding as well as Supreme Court precedent. As a matter of text and history, the focus of the Due Process Clause was a protection against the arbitrary use of the “exclusive sovereign prerogative to coerce or restrain action.” See Matthew Pritchard, Reviving DeShaney: State-Created Dangers and Due Process First Principles, 74 Rutgers U. L. Rev. 161, 192 (2021). The notion of an affirmative duty to help comes nowhere from our history or the meaning of the text. And under Court precedent, the Clause is “not . . . a guarantee of certain minimal levels of safety and security.” DeShaney, 489 U.S. at 195. The Clause was meant “to protect the people from the State, not to ensure that the State protected them from each other.” Id. at 196. And so “a State’s failure to protect an individual against private violence simply does not constitute a violation of the Due Process Clause.” Id. at 197. Under the proper understanding of the Due Process Clause, outside the custodial context, the government bears no duty to act to protect or help others—even faced with devastating and tragic actions by private actors. And combining non- particularized affirmative acts with non-affirmative acts doesn’t save the majority’s novel construction of the Fourteenth Amendment. The implications of this ruling are again far-reaching. Under the majority’s reasoning, under some circumstances, state officials now bear an affirmative constitutional obligation to ensure the safety and security of any civilian who encounters a violent or dangerous private actor. It’s hard to see how this isn’t a transformative reading of the Due

ESTATE OF SOAKAI V. ABDELAZIZ 45

Process Clause. Now, instead of government action being the basis for a constitutional claim, inaction may violate due process. The majority tries to downplay the significance of its ruling by claiming that it doesn’t apply to “ordinary failure to render aid.” Maj. Op. 29. But that is no solace— and it only adds confusion to a confusing doctrine. And, unsurprisingly, as a product of the majority’s judicial innovation, this new interpretation of the state- created-danger doctrine was not clearly established law. The majority all but concedes this. See Maj. Op. 26 (“[N]or are we aware of . . . a controlling case applying the state-created danger theory to injuries suffered by bystanders after a high- speed police chase.”). Indeed, the majority doesn’t even try to suggest that its creative approach to manufacturing “affirmative conduct” through a combination of non- particularized affirmative conduct and inaction is clearly established law. But the majority presses ahead based on the broadest of generalities. It proclaims it “established that officers violate substantive due process when they affirmatively place an individual in danger and then . . . do nothing to address that danger.” Id. at 26. Once again, we ignore the Supreme Court’s “repeated[]” admonition to “the Ninth Circuit in particular” “not to define clearly established law at a high level of generality.” Ashcroft v. al-Kidd, 563

U.S. 731, 742 (2011).

The majority seeks to satisfy the clearly established law prong by invoking four cases: Wood v. Ostrander, 879 F.2d 583 (9th Cir. 1989), Bracken v. Okura, 869 F.3d 771 (9th Cir. 2017), Maxwell v. Cnty. of San Diego, 708 F.3d 1075 (9th Cir. 2013), and Penilla v. City of Huntington Park, 115 F.3d 707 (9th Cir. 1997). None of these decisions place the Officers’ duty to summon or render medical assistance “beyond debate.” See Kisela v. Hughes, 584 U.S. 100, 104 (2018). First, Wood only ruled that an officer who affirmatively “abandon[ed]” a woman on the side of the road in a high- crime area in the middle of the night violated the state- created-danger doctrine. 879 F.2d at 592. There, the officer’s actions were affirmative conduct—the plaintiff asked the officer how she would get home after the officer arrested the driver of her car for a DUI. Id. at 586. In response, the officer simply told her she needed to get out of the car and then left her on the side of the road. Id. She was later raped that night. Id. According to our court, the Due Process Clause was implicated because of the “police officer’s roadside abandonment of non-arrested third parties.” Id. at 592. We then held that an officer “stranding a lone woman in a high-crime area at 2:30 a.m.” was a constitutional violation. Id. So Wood is about a police officer’s actions—and it was a due process violation for an officer to “abandon[] passengers of arrested drivers under circumstances which expose[d] them to unreasonable danger.” Id. at 593 (emphasis added). Thus, Wood doesn’t put the Officers on notice that they must affirmatively call for or render medical assistance. Wood only says that officers can’t affirmatively abandon civilians in dangerous situations. Second, Bracken embraced a state-created-danger claim when an off-duty police officer moonlighting as a hotel security guard affirmatively participated in the beating of a man at a hotel party by other security guards. 869 F.3d at 775. During the incident, the off-duty officer warned the man that he was trespassing and asked for his identification. Id. at 779. After being “jostl[ed]” and “yell[ed]” at, the man repeatedly asked if he could leave, but the officer “asserted

ESTATE OF SOAKAI V. ABDELAZIZ 47

his authority over” the man, continued to ask for his identification, and blocked his exit. Id. Soon after, other hotel security guards arrived and beat the man unconscious. Id. at 775. We held that the officer was not entitled to qualified immunity because the officer “affirmatively prevented [the man] from leaving the party and ensured that [the man] remained under the control of the hotel’s security guards.” Id. at 779. In other words, even though he didn’t throw any punches, the officer “was still an active participant in the incident.” Id. If the officer “let him leave,” the man would have left uninjured. Id. Again, this case provides no notice to the Officers here. By failing to summon assistance, they didn’t block or otherwise interfere with medical services. Nor did they affirmatively exacerbate their injuries by doing nothing. Third, Maxwell allowed a state-created-danger claim after police officers affirmatively “refused to let [an] ambulance leave immediately” with a gunshot victim because the officers thought the victim needed to be interviewed first. The victim died en route even though her injuries were survivable. 708 F.3d at 1081. We concluded that “[i]mpeding access to medical care” was sufficient government affirmative action to sustain a due process claim. Id. at 1082. Indeed, the evidence showed that officers “affirmatively increased that danger [to the gunshot victim] by preventing her ambulance from leaving.” Id. Thus, the Maxwell affirmative conduct was obvious—impeding an ambulance from leaving. Here, we have nothing of the sort. At most, the Officers failed to call an ambulance—that’s very different than stopping one. And finally, Penilla affirmed the denial of qualified immunity after police officers affirmatively “made it impossible for anyone to provide emergency medical care to” a victim in medical distress. 115 F.3d at 710. In that case, police officers responded to a 911 call by neighbors and a passerby about a man in “grave need” of medical care. Id. at 708. Rather than permitting medical help, the officers affirmatively cancelled the request for paramedics. Id. at 710. They then dragged the man from his porch in public view and left him in an empty house alone and locked the door behind them. Id. The man was found dead the next day. Id. Again, the officers engaged in several affirmative steps that left the man in a place where he could receive no medical attention. If they simply did nothing, the man would have received the help he needed. The Officers’ conduct here bears no resemblance to these many affirmative steps. The Officers neither interfered with Plaintiffs receiving medical attention nor placed them in a position rendering medical attention impossible. So in these cases, the state actors affirmatively abandoned a plaintiff in a dangerous situation, affirmatively prevented a plaintiff from leaving a dangerous situation, affirmatively impeded medical assistance from helping the plaintiff, or affirmatively made it impossible to assist the plaintiff. None of those situations would put Officers on notice of the alleged constitutional violation here—that they did nothing after a tragic accident. There’s simply no clearly established law here.

III.

For the reasons discussed above, the Officers were entitled to qualified immunity, and we should have reversed the district court’s order. I respectfully dissent.