Hughes et al v. Apple, Inc.

Filing 73

Order by Judge Vince Chhabria granting in part and denying in part the 50 Motion to Dismiss.(vclc3, COURT STAFF) (Filed on 3/15/2024)

UNITED STATES DISTRICT COURT
NORTHERN DISTRICT OF CALIFORNIA

LAUREN HUGHES, et al.,

Case No. 22-cv-07668-VC

Plaintiffs,
ORDER PARTIALLY DENYING AND
PARTIALLY GRANTING THE
MOTION TO DISMISS

v.
APPLE, INC.,
Defendant.

Re: Dkt. No. 50

The AirTag is a small, affordable location-tracking device manufactured and sold by
Apple. It is marketed as a convenient way to keep track of personal items, like car keys. The
AirTag reports its location to its owner’s cell phone, so you can simply attach the AirTag to your
car keys and then you can use your cell phone to find them. But people do not always use a
product as marketed. The thirty-eight plaintiffs in this case allege they were stalked by someone
who used an AirTag to follow their location and movements.
From the beginning, it was obvious that the AirTag would be an especially useful tool for
stalkers. So Apple has designed it with features that aim to diminish the ability of stalkers to use
it effectively. But the plaintiffs allege that those features are inadequate, and that Apple could
and should have done more. They have filed a 131-page class action complaint against Apple,
asserting many different legal claims based on Apple’s alleged failure to mitigate the dangers
created by the AirTag.
Most of the claims are inadequately pled, and they are dismissed in a separate ruling. But
three of the plaintiffs have stated claims for negligence and strict products liability under
California law. Those plaintiffs allege that, when they were stalked, the problems with the

AirTag’s safety features were substantial, and that those safety defects caused their injuries. This
ruling explains why those claims survive, even though it’s a close question. Apple may
ultimately be right that California law did not require it to do more to diminish the ability of
stalkers to use AirTags effectively, but that determination cannot be made at this early stage.
I
A
This subsection describes and quotes the allegations that two of the California plaintiffs
make about their experiences being stalked through the use of AirTags. These allegations must
be treated as true for purposes of this motion to dismiss.
On Father’s Day weekend in 2022, John Kirkman’s estranged wife showed up at his
house, banged on the door, rang the doorbell, drove up and down the street honking her horn,
and did not leave for several hours. Kirkman had noticed “signs of aggressive and erratic
behavior” during the marriage, which made him fear for his safety. Because of that, Kirkman
made efforts to keep his estranged wife from discovering where he was living. But she found his
new home, and her behavior that day scared and upset both him and his daughter, who was also
in the house at the time.
When Kirkman was driving home from work the next day, he saw a notification on his
iPad: it informed him that “an unknown AirTag was detected.” Since he did not know what that
meant, Kirkman looked online “to figure out what an AirTag was and how to locate it.” He
learned that his daughter had previously received a similar alert when riding in the car with him,
but she did not bring it up because she did not know what it meant. Kirkman spent several days
searching for the AirTag. “Ultimately, he found one duct-taped to the inside of the rear bumper
cover of his vehicle.”
Kirkman remains afraid of future AirTag stalking by his estranged wife. He called the
police and filed an online report for harassment, but “nothing ultimately came of that.” Since the
incident, he has “suffered from anxiety,” is “unable to relax,” feels “constantly on edge,” and has
“a great deal of difficulty sleeping.” He feels the need to remain constantly aware of his

2

surroundings and to stay up late into the night, to make sure he is not being stalked. He is afraid
to leave his home and has difficulty engaging in activities that he previously enjoyed. Put simply,
“he doesn’t feel like the same person he was a year ago.”
In 2021, Àine O’Neill moved to Los Angeles to pursue a career in film and television.
She entered the United States on an O-1 artist visa, which took her two years to acquire. After
her move, she found a job in the industry, took acting classes, did theater in the park with friends,
played tennis, did stand-up, joined a screenwriting group, and enjoyed an active social life.
In late November 2022, O’Neill saw a notification on her phone about an AirTag. She
searched her purse and car but did not see one. Then she “tried to get the AirTag to make a
sound.” Upon hearing a sound coming from her car, but not from inside the car, she concluded
the AirTag was likely underneath the vehicle. Then she and her roommate both searched for the
device, but neither could track it down. Although O’Neill could not locate the AirTag, her phone
showed her everything the AirTag owner could see about her: “where she worked, where she
parked her car, her home address, and everywhere she had been.” So she “texted with Apple,”
and Apple “confirmed the existence of the AirTag but informed her that they were unable to
disable [the] tag without physical possession.”
O’Neill called the police the next morning and made a report. But the police told her that,
even though “there’s no way that this is anything other than sinister,” they “could not help her
fully until she had the physical AirTag in hand.” O’Neill then “took her car to Pep Boys” for
help, but they were also unable to find the AirTag. Then she took the car to another auto shop,
which told her that they would have to strip the car to find it. Because she could not afford the
price they quoted her, she left.
The stalking destroyed O’Neill’s life in Los Angeles. The week after she found the
AirTag, she “called in sick to work,” “canceled her acting classes and all other plans,” and
“found herself barely able to get of bed.” O’Neill “sunk into a deep depression.” And she was
terrified, “peering out through her blinds at all hours,” “leaving lights on in her apartment
throughout the night,” and having panic attacks whenever she was alone in the apartment. The

3

phone alerts continued for more than a month. O’Neill decided that “she could no longer live in
California, as she had no way of revealing the identity of her stalker or properly gauging the
level of danger she was in.”
So O’Neill moved back to Ireland. She gave up a recently booked recurring acting role on
a television series, the money she had been saving to buy a home but was forced to spend on an
emergency move, the visa that she had worked so hard for and that she still had a year left on
(with plans to try to renew it), and the rest of the professional and personal life she had spent two
years building in California. Back in Ireland, she still suffers as a result of the stalking: she feels
embarrassed, humiliated, devastated, and afraid.
Kirkman and O’Neill are part of a group of thirty-eight plaintiffs who were stalked with
AirTags. Three additional plaintiffs were stalked in California; the rest are scattered across the
United States and Canada. The experiences vary, but each plaintiff alleges that their stalker used
an AirTag, and that the involvement of the AirTag caused them harm.
B
The AirTag is a small tracking device manufactured and sold by Apple. The user can
follow the location of their AirTag through their iPhone. The purpose of the product is to help
people keep track of things like keys and purses—if you have an AirTag on your keychain or in
your purse, you’ll be able to find those items using your phone (which, of course, never leaves
your side).
But the plaintiffs allege that the AirTag has also “revolutionized the scope, breadth, and
ease of location-based stalking.” They point to several features that differentiate the AirTag from
other tracking devices on the market: its accuracy, its ease of use, and its affordability. The
AirTag allegedly operates via Bluetooth—it “emits signals that are detected by Bluetooth
sensors” on other Apple products, and those Apple products then report the AirTag’s location.
Bluetooth range is approximately thirty feet. And the plaintiffs allege that, in much of United
States, “one is never more than 100 yards away from an Apple device.” Thus, the AirTag can
transmit accurate location data from just about any populated area. Apple product users may be

4

familiar with this Bluetooth sensor system from the “FindMy” network, which Apple also uses to
locate devices like iPhones and iPads (either to enable people to share their location with their
friends and family, or to find their missing devices). The AirTag works the same way. That leads
to the second alleged differentiating feature: the AirTag is easy to use because it fits “seamlessly
into Apple’s existing suite of products.” Finally, the plaintiffs allege that it is affordable, only
twenty-nine dollars per device, and small, about the size of a quarter.
Apple released the AirTag in April 2021. The plaintiffs allege that in the run-up to and
immediate aftermath of the launch, “advocates and technologists urged the company to rethink
the product and consider its inevitable use in stalking.” But, the plaintiffs assert, Apple ignored
these warnings. When it released the product, Apple touted several “mitigation features that it
claimed rendered the devices ‘stalker proof.’” And the plaintiffs acknowledge that Apple has
implemented a number of safety features aimed at the risk of AirTag stalking—some were
operational when the product launched, and some have become newly available or have been
improved since launch. But the plaintiffs allege that these safety features are deficient.
Take, for example, the screen alerts. If a device that runs Apple’s operating architecture
(iOS) “detects an unknown AirTag moving with the device,” then the screen will display an
alert. The alert is a pop-up text notification, and it is allegedly how Kirkman, his daughter, and
O’Neill learned that they were being stalked with an AirTag. The recipient of the alert can also
make the AirTag play a sound. And, based on O’Neill’s account, it seems that the notification
also enables the recipient to see what the AirTag owner can see—but it is not entirely clear from
the complaint how this feature works.1
1

It is worth noting that the allegations leave a lot of uncertainty about potentially relevant
aspects of the product. The screen alerts are one example. What text is displayed to the user, and
has that changed over time? What can a user see and do if they click on the alert, and has that
changed over time? If a user is able to see what the AirTag has tracked, how far back are they
able to look? The plaintiffs’ general allegations, combined with the stories of individual
plaintiffs, also raise some inconsistencies. That might be a result of imprecise pleading. It might
be a result of the features working inconsistently. And it might also be a result of Apple making
and testing changes to the safety features. Ultimately, a clearer timeline will be necessary for
multiple aspects of the plaintiffs’ claims (e.g., when was Apple in breach of a duty to use due
care, which features were available when a particular plaintiff was injured, etc.). But, as

5

The plaintiffs argue that there are numerous issues with these screen alerts. To start, there
is allegedly significant delay between the moment the AirTag is detected traveling with the
device and the moment the alert shows up. “[T]he timing of the alerts can vary depending on the
iPhone’s operating system and location settings”—some users report up to a day of tracking
before receiving the alert, while Apple “estimates it takes between four and eight hours.” And
that timeframe is apparently an improvement. The plaintiffs allege that when the AirTag was
first released, the notification would not be sent until seventy-two hours after detection. On top
of that, many of the plaintiffs report not understanding what the notifications meant when they
saw them, suggesting that the phrasing may be confusing or unclear. Additionally, people
allegedly cannot trigger a search for an unknown AirTag themselves: they must wait for Apple’s
system to send an alert. Thus, if the recipient does not immediately act on the notification—
either because they do not know what it means or because they are unable to do so at the time
they receive it—and it goes away, then the user cannot know when or whether they will receive
another notification. For example, one plaintiff alleges that her daughter got a screen alert while
they were driving, but by the time they had pulled off the road to search for the device, the alert
was gone. Beyond that, the plaintiffs assert that not all Apple product owners can receive these
alerts. They are only available for people with iPhones, iPads, or iPod Touches that run iOS
version 14.5 or later.2 Also, if a user disables “Location Services” in their device settings, then
the user will not be able to receive the Unknown AirTag screen alerts—but this “consequence” is
allegedly not explained to users.
Another feature that the plaintiffs take issue with is the sound alert. The plaintiffs allege
that an AirTag will play a “chime-like sound” when it is away from its owner for too long.
Again, the noises do not begin immediately—Apple allegedly says they will play within eight
discussed below, for certain claims asserted by these California plaintiffs, the complaint has
alleged enough to survive the motion to dismiss.
2
Apple has a different device-based system for alerting Android users. The complaint also
alleges serious defects with that system. But since this opinion addresses only the California
plaintiffs, and all of them are identified as iOS users in the complaint, those allegations are not
presently relevant.

6

and twenty-four hours of separation from the AirTag’s owner. And it’s unclear how long the
noise will continue, or if it starts and stops in intervals. But the primary issue is with the sound
itself: its volume and its distinctiveness. The sound is allegedly around sixty decibels, which the
plaintiffs describe as “approximately as loud as a normal conversation between two people, or
background music.” The plaintiffs allege that “the sound is not particularly distinctive,” quoting
a reporter who wrote that “the sound was easy to confuse with all the other beeps and boops
gadgets make these days.” These issues with the sound are particularly problematic, the plaintiffs
assert, “if the victim is hearing impaired or in a loud environment, or if the stalker places the
AirTag in a place where it will be muffled or out of range of hearing (like the outside of a car).”
On top of that, people have apparently discovered how to disable the speaker. The plaintiffs
allege that tutorials for deactivating or removing the speaker are easily accessible online, and that
modified “silent AirTags” are available for purchase on mainstream e-commerce websites.
The plaintiffs also complain about how difficult it is to deactivate AirTags, alleging they
can be deactivated only manually, by removing the battery. This obviously requires physical
possession of the actual device. So the stalking victim has to actually locate the unknown
AirTag—otherwise, the stalking can simply continue. And deactivation has other downsides.
Removing the battery allegedly sends an alert to the AirTag’s owner, so they immediately know
that the device has been found. Also, the plaintiffs allege that law enforcement agencies have
cautioned that tampering with the battery can contaminate the AirTag as evidence.
The plaintiffs argue that these defects render the product too unsafe. The lawsuit includes
many different claims and many different theories of liability, but the two strongest ones are
common law tort claims for negligence and strict products liability. There is no dispute that the
tort claims brought by people who were victims of stalking in California are governed by
California tort law. This opinion will discuss the viability of those tort claims. As noted above, a
separate ruling addresses the other causes of actions and the plaintiffs who were stalked outside
of California.

7

II
The plaintiffs assert that Apple owes AirTag-stalking victims “a duty of care in its
design, marketing, and introduction into the market of its AirTags.” They allege that Apple
breached this duty by “rushing AirTags to market with insufficient safeguards to prohibit their
use for stalking purposes.” And they allege this caused their injuries. In other words, Apple
allegedly breached its duty of care to the plaintiffs by “putting them at an unreasonable risk of
harm through the reasonably foreseeable actions of third-party stalkers.” Apple contests both the
duty and causation elements of the plaintiffs’ negligence claims.
A
The baseline rule in California is that everyone has a duty “to exercise, in his or her
activities, reasonable care for the safety of others.” Kesner v. Superior Court, 1 Cal.5th 1132,
1142 (2016) (quoting Cal. Civ. Code § 1714(a)); see also Modisette v. Apple Inc., 30
Cal.App.5th 136, 143 (2018). But courts can carve out exceptions to the “broad principle enacted
by the Legislature that one’s failure to exercise ordinary care incurs liability for all the harms that
result.” Kesner, 1 Cal.5th at 1143. The California Supreme Court has emphasized that the
question for courts is not whether to create a new duty, as Apple characterizes it, but rather,
whether to carve off an exception to the baseline duty that applies to everyone. Id. And courts
should create such an exception only when it is “clearly supported by public policy” and when it
is possible to “promulgate relatively clear, categorical, bright-line rules of law applicable to
general classes of cases.” Id. at 1143–44. For that reason, the analysis of duty “occurs at a higher
level of generality” than the facts of each individual case. Id. at 1144. In other words, the duty
inquiry examines categories of conduct and categories of potential plaintiffs who allege certain
categories of injury. For example, the duties of a landowner to a trespasser. Then the other
elements of the tort claim—breach, causation, and injury—must be determined based on the
particular facts of the individual case. In other words, whether a particular defendant breached
their duty to a particular plaintiff, if that breach caused that plaintiff’s harm, and what damages
the plaintiff suffered.

8

The factors that are meant to guide courts in determining whether an exception is
appropriate are called the Rowland factors, based on the California Supreme Court’s discussion
of them in Rowland v. Christian, 69 Cal.2d 108, 112–13 (1968). The factors are overlapping and
vague. Courts often sort them into two different categories based on their focus, with “the first
addressing foreseeability and related concepts and the second addressing public policy
considerations.” Social Media Cases, Nos. JCCP5255, 22STCV21355, 2023 WL 6847378, at
*24 (Cal. Super. Ct. Oct. 13, 2023). The foreseeability factors are “the foreseeability of harm to
the plaintiff, the degree of certainty that the plaintiff suffered injury, [and] the closeness of the
connection between the defendant’s conduct and the injury suffered.” Id. (quoting Rowland, 69
Cal.2d at 113). But even if the plaintiff prevails on the foreseeability factors, courts will still
sometimes find that there is no duty if the policy factors counsel otherwise. Id. at *26. The policy
factors are “the moral blame attached to the defendant’s conduct, the policy of preventing future
harm, the extent of the burden to the defendant and consequences to the community of imposing
a duty to exercise care with resulting liability for breach, and the availability, cost, and
prevalence of insurance for the risk involved.” Id. at *24 (quoting Rowland, 69 Cal.2d at 113).
As noted, the analysis of these factors does not occur at the level of the individual incident. So,
as applied to this case, the analysis is about whether the manufacturer of a tracking device should
be excused from the duty to use reasonable care in the creation, design, and marketing of that
device to prevent or mitigate harm from stalking.
As an initial matter, Apple contests the applicability of this framework to the plaintiffs’
claims. It argues that the California statutory provision creating the baseline rule imposing a duty
of care does not govern the asserted defects with the AirTags. The provision states, in relevant
part: “Everyone is responsible . . . for an injury occasioned to another by his or her want of
ordinary skill in the management of his or her property or person.” Cal. Civ. Code § 1714(a).
Apple argues that by selling the AirTags to the stalkers, it stepped outside of the provision’s
scope, because the AirTags were no longer Apple’s property. But this argument conflicts with
the case that Apple relies on most heavily: Modisette v. Apple, 30 Cal. App. 5th 136 (2018).

9

There, the California Court of Appeal ruled that Apple owed no duty to use reasonable care to
prevent harm from car accidents caused by drivers who were distracted by their iPhones while
driving. This was, of course, a favorable ruling for Apple. But the court reached its ruling by
applying the Rowland factors to create an exception to section 1714(a)’s general rule; it did not
rule that section 1714(a) becomes inapplicable once a company sells its product to a third party.
See id. at 143–52. Absent strong reason to believe that the California Supreme Court would
apply a different framework to the scenario presented here and in Modisette, this framework
controls. See In re Watts, 298 F.3d 1077, 1082–83 (9th Cir. 2002). Apple has not identified case
law or other evidence that would suggest a different outcome on this issue at the California
Supreme Court.3 The question, then, is not whether Modisette’s framework applies. It is whether
the Rowland factors absolve Apple of a duty to use reasonable care in its AirTag design to
mitigate harm from stalking in the same way they absolved Apple of a duty to use reasonable
care in its iPhone design to mitigate harm from distracted driving.
Foreseeability of harm to the plaintiff. The foreseeability of the harm is often described
as “the most important factor” in the analysis of whether a duty exists. Social Media Cases, 2023
WL 6847378, at *24 (quoting Kesner, 1 Cal. 5th at 1145)). Common sense alone compels the
conclusion that harm from stalking is a foreseeable consequence of making and selling a tracking
device, especially a small, affordable, consumer-friendly tracking device. Similarly, the court in
Modisette concluded that harm from car accidents caused by cell phone use by drivers was
sufficiently foreseeable to support a duty. Modisette, 30 Cal.App.5th at 144.
In analyzing this factor, the Modisette court also considered evidence that Apple itself

Apple’s argument is based on language in some case law stating that there is “no duty to
control the conduct of another.” Tarasoff v. Regents of University of California, 17 Cal.3d 425,
435 (1976); see also In re Deep Vein Thrombosis, 356 F.Supp.2d 1055, 1066 (N.D. Cal. 2005)
(citing In re Firearm Cases, 126 Cal.App.4th 959 (2005)). But that is not the plaintiffs’ claim
here. The plaintiffs are not asserting that that Apple has “a duty to initiate an affirmative program
of investigation and sanctioning of wayward” users. In re Firearm Cases, 126 Cal.App.4th at
973 n.12. Nor that Apple ought to “control” the stalkers themselves. Rather, the plaintiffs are
asserting that Apple has a baseline “duty to use ordinary care in the conduct of [its] business” to
avoid causing harm to others. Id.
3

10

was aware of the risk of harm of distracted driving. Id. at 143–44. While that case-specific
inquiry seems out of step with the instruction to assess the existence of a duty at a higher level of
generality, the plaintiffs here would also prevail on the specifics. The plaintiffs allege not only
that Apple should have known about the risks of AirTag stalking but also that it actually did
know. The plaintiffs allege both pre- and post-launch warnings from “advocates and
technologists” about the dangers its product posed for stalking victims. Contemporaneous news
articles suggested that Apple executives were touting the product as “stalker proof.” Then, after
the product launched, the plaintiffs allege that many articles describing the ease of stalking with
AirTags were featured in national news outlets, as were multiple stories of actual AirTag
stalking.
Apple asserts that the “specific injuries” that the plaintiffs suffered were not foreseeable.
Dkt. No. 53 at 5. Relatedly, Apple argues that, since it implemented some safety measures, the
Court should be asking whether stalking despite the safety measures was foreseeable. See, e.g.,
Dkt. No. 62, Hearing Transcript at 6:16–17:9. But the duty question is more general than that—it
is whether the “category of negligent conduct at issue is sufficiently likely to result in the kind of
harm experienced.” Kesner, 1 Cal.5th at 1145. Whether Apple’s conduct renders it liable for
specific injuries experienced by certain stalking victims is an individual, case-level question.
Apple may ultimately be right that some of the thirty-eight plaintiffs experienced unforeseeable
harms (certain actions by the stalker, perhaps, or damage to property). But that goes to proximate
cause, not duty. Similarly, Apple may ultimately be right that it took due care in making the
product safe for stalking victims. But that goes to breach, not duty. It is enough to say that severe
emotional distress as a result of being stalked by AirTags was foreseeable.
Certainty of injury. This factor includes considerations about “the degree of certainty that
the plaintiff suffered injury” and “concerns about the existence of a remedy.” Kesner, 1 Cal.5th
at 1148. The plaintiffs allege they suffered emotional distress from the AirTag stalking. Apple
responds that this factor can support the existence of a duty only where the injuries alleged are
physical. It invokes a footnote from Modisette that commented that “the certainty of injury factor

11

has been noted primarily, if not exclusively, when the only claimed injury is an intangible harm
such as emotional distress.” Modisette, 30 Cal. App. 5th at 145 n.7 (quoting Kesner, 1 Cal. 5th at
1148). But all this means is that the certainty-of-injury factor doesn’t come up when the alleged
injury is physical; it tends to come into play only when the alleged injury is emotional. Certainly,
there are some situations where it will be difficult to assess whether an alleged emotional injury
is genuine. See Billy v. Arthur Young & Co., 3 Cal.4th 370, 421 (1992) (Kennard, J., dissenting).
And in those situations, the certainty-of-injury factor can cut against finding a duty, as courts are
concerned about opening the floodgates to insincere claims, and thus overdeterring socially
useful conduct. But serious emotional harm is a likely result of being stalked. So the risk of an
influx of unbothered stalking victims seeking tort recovery seems fairly low. Overall, this factor
does not favor either side.
Closeness of the connection between the defendant’s conduct and the injury suffered.
This factor assesses how close the connection was between the defendant’s conduct and the
plaintiff’s injuries. Case law is clear that the involvement of a third party in causing the injury
does not preclude finding a duty of care: “One’s general duty to exercise due care includes the
duty not to place another person in a situation in which the other person is exposed to an
unreasonable risk of harm through the reasonably foreseeable conduct . . . of a third person.”
Modisette, 30 Cal. App. 5th at 145 (quoting Kesner, 1 Cal. 5th at 1148). But the third party’s
involvement is one layer of attenuation between the conduct and the injury. So, in applying this
factor, courts will consider how many other layers of attenuation are present.
In one case, the California Supreme Court held that a company had a duty to family
members and cohabitants of employees who were exposed to asbestos when the employees
returned home. Kesner, 1 Cal.5th at 1156. In another case, the California Supreme Court found a
duty where “a police officer motioned a speeding motorist to stop in the center median of a
highway,” at which point the motorist’s car was struck by a third party, seriously injuring the
passengers. Modisette, 30 Cal. App. 5th at 146 (citing Lugtu v. California Highway Patrol, 26
Cal.4th 703 (2001)). But the court in Modisette found that this factor counseled against a duty.

12

Unlike the employer’s use of the asbestos, and the cop telling the driver to pull over, Apple’s
design of the iPhone “did not put the danger in play”—rather, the design “simply made [the third
party’s] use of the phone while driving possible, as does the creator of any product (such as a
map, a radio, a hot cup of coffee, or makeup) that could foreseeably distract a driver using the
product while driving.” Id. Apple argues that the same logic applies in this case; the AirTag’s
design simply made the stalking possible—it did not put the danger of stalking into play.
There is certainly real attenuation here, and this factor does cut somewhat against finding
a duty, but not to the extent that Apple tries to claim. For one, the plaintiffs have detailed factual
allegations about harm from being stalked with a location-tracking device specifically, as
opposed to in person or over social media. They allege fear of driving, not thinking anywhere is
safe, and not knowing how to protect themselves. A car accident caused by someone looking at a
phone is roughly the same as a car accident caused by someone looking at a map. But according
to the allegations in this case, the product causes a particularly devastating harm. For another, the
connection between the purpose of the product and the harm to the third parties is closer than in
Modisette. A smartphone serves a great multitude of functions, one of which is video calls, and
it’s only incidental to any of those functions that a user would be distracted and crash their car.
But the sole function of a tracking device is to track location—and the injuries arise directly from
the tracking device performing that function. So it is not obvious that Apple did not, in a sense,
“put the danger in play.”
Additionally, there are fewer links in the chain of causation in this case, as compared to
Modisette. There, the driver needed to choose to look at his iPhone, he needed to choose to use
FaceTime, he needed to be driving in a manner that caused him to fail to see a hazard on the
road, he needed to be driving such that he would be in an accident because of that hazard, the
other party needed to be unable to avoid the accident, and then there needed to be injuries
resulting from the accident. Here, it is simpler. Apple put out a product that would be used for
stalking, and it was in fact used for stalking. The stalker has to choose to use the AirTag for
stalking, and then injury results. Thus, overall, this factor weighs against finding a duty, but far

13

less strongly than in Modisette.
Considered altogether, the foreseeability-related factors weigh against creating an
exception to the general duty rule. Next come the public policy factors.
Moral blame attached to the defendant’s conduct. California courts have “previously
assigned moral blame, and [ ] have relied in part on that blame in finding a duty, in instances
where the plaintiffs are particularly powerless or unsophisticated compared to the defendants or
where the defendants exercised greater control over the risks at issue.” Kesner, 1 Cal.5th at 1151.
Both are true here. Victims of stalking have little to no power over the mechanisms their stalkers
use. But the maker and seller of a product that is particularly useful to stalkers does have power
over that product. And while stalking victims are not unsophisticated as a general matter, the
public is largely unsophisticated when it comes to how tracking devices work and what they can
do to protect themselves. Moreover, in assessing this factor, courts have also considered whether
the defendant “benefitted financially” from its conduct. Id. And Apple profits from all sales of
AirTags—whether the sale is to someone who later uses it to track their car keys, or to track a
human being without their consent. Thus, this factor weighs in favor of finding a duty.
Policy of preventing future harm. The existence of a duty is certainly supported by the
state’s policy of preventing future harm. “The overall policy of preventing future harm is
ordinarily served, in tort law, by imposing the costs of negligent conduct upon those
responsible.” Kesner, 1 Cal.5th at 1150. However, “for a category of negligent conduct,” that
overall policy may be “outweighed” if there are “laws or mores indicating approval of the
conduct.” Id. Apple does not note any laws or mores in support of its conduct. And there is good
reason to believe that imposing liability for negligence in the design, marketing, and sale of
location trackers would prevent future harm—the manufacturers of tracking devices are in the
best position to engage in conduct that promotes the product’s safe use. Cf. id. (considering
whether there is a future benefit to imposing liability). Similarly, in Modisette, the court
concluded that the policy of preventing future harm supported the (ultimately unsuccessful)
argument that Apple had a duty in the context of iPhones distracting drivers. 30 Cal. App. 5th at

14

145.
Extent of the burden to the defendant and consequences to the community of imposing a
duty to exercise care with the resulting liability for breach. This factor, which considers the
burden on the defendants and broader society of imposing a duty to exercise due care, is perhaps
the most important of the policy factors. See Kesner, 1 Cal.5th at 1144. It was a key reason the
Modisette court found an exception to the duty rule. The Modisette court ultimately concluded
that putting the burden on cell phone manufacturers to design phones to prevent or reduce
distraction while driving would go too far. The court quoted from another opinion, which
compared the phone manufacturers to a “cosmetic manufacturer or map designer,” suggesting
that a duty in this case would extend similarly to the makers of all products that might be
distracting to a driver. Modisette, 30 Cal. App. 5th at 149. The Modisette court also looked at
state policy, finding it relevant that California permitted drivers to use cell phones in the car
under certain circumstances, and even encouraged drivers to use their cell phones while driving
to do things like report accidents to the authorities. Id. at 150. Finally, the Modisette court
emphasized just how prevalent cell phones are in modern society, quoting the United States
Supreme Court’s observation that cell phones are “such a pervasive and insistent part of daily
life that the proverbial visitor from Mars might conclude they were an important feature of
human anatomy.” Id. (quoting Riley v. California, 573 U.S. 373, 385 (2014)).
But each of those considerations from Modisette comes out differently in this case. First,
there is no similar spillover effect to other kinds of products (as with other driving distractions),
and there is no similar precedent for rejecting a duty of care for products that are effective tools
for stalking (as with the cases rejecting a duty for distracting drivers). Second, there is no state
policy permitting or encouraging, in certain circumstances, individuals to track other people’s
locations without their consent. Stalking is, of course, illegal. The plaintiffs have also pointed to
laws emphasizing California’s general policy commitments to personal privacy. Apple could
note that California has not chosen to ban tracking devices, which have existed for a long time,
nor has it barred AirTag itself, which has been on the market for a few years now. But that

15

absence of action is a far cry from the affirmative policy statements that the court found
meaningful in Modisette. Finally, Bluetooth trackers are not nearly as central to everyday life as
cell phones. Apple emphasizes the useful, legal functions that AirTags do serve. Dkt. No. 62,
Hearing Transcript at 28:23–29:1. There is no dispute that the AirTag is a useful tool for keeping
track of items, but its societal value obviously pales in comparison to that of a cell phone.
At the hearing, Apple urged the Court to take an even broader view, looking at the
societal importance not only of the tracking chip itself but also of the entire “FindMy” network.
Id. at 29:7–15; 29:25–30:14. Apple asserted that the FindMy network—which is the Bluetooth
network generated by iPhones and iPads that the AirTags operate by tapping into—provides
people with the ability to see whether their “child is going to school” or whether their “family
members are safe.” Id. at 30:1–3. Just as the Modisette court looked beyond Facetime to the cell
phone, Apple urges that this Court look beyond the tracking device that caused the injuries in
these cases to the entire location-tracking “ecosystem.” Id. at 30:9. But that broader lens was
made appropriate by the fact that the duty that the plaintiffs hoped to impose in Modisette would
have broadly burdened cell phone manufacturers. There is no similar theory for why imposing a
duty on the maker of a Bluetooth location-tracking device would burden Apple’s overall
location-tracking system generated by its other, multifunctional products. Based on the
description of the technology in the complaint, Apple could cease manufacturing AirTags today
and the rest of the FindMy system would still function just the same.4
In the end, most of Apple’s arguments for why it does not have a duty to use due care
boil down to different versions of “but we implemented sufficient safety measures in the
AirTags.” It may ultimately be that Apple will not bear liability for any of these injuries, either
because it did in fact take due care in designing the AirTag, or because its safety measures were
effective enough that the product design cannot be said to have caused the injuries. But that
4

Apple may have meant that a duty to take due care to prevent stalking would burden them not
only with regard to the AirTag product but also with regard to other Apple products—products
that also perform location tracking. But this ruling is not about those products. And all of the
Rowland factors, not just this one, could come out differently as applied to a different product.

16

misunderstands the duty inquiry. If a party could avoid a finding of a duty of care by asserting
that it did indeed take due care, the whole exercise would break down. The question at the duty
stage is whether it’s appropriate to excuse a party engaged in a certain category of conduct—the
making, marketing, and selling of a location-tracking device—from having a baseline duty to use
reasonable care to avoid causing a certain kind of injury—emotional distress and other harms
resulting from stalking with the device. Whether each individual case in that category will meet
the rest of the elements for a negligence claim is a different question.5
B
Apple also argues that the plaintiffs do not adequately allege that the purported breach,
putting AirTags on the market without sufficient safety mechanisms, proximately caused the
injuries experienced by the stalking victims.
It is certainly true that the intentional acts of third-party wrongdoers can cut off the chain
of causation. However, “a third party’s conduct is a superseding force cutting off the defendant’s
liability only if it was unforeseeable and the injury it caused was unforeseeable.” Huang v. The
Bicycle Casino, Inc., 4 Cal. App. 5th 329, 348 (2016); see also Soule v. General Motors Corp., 8
Cal.4th 548, 573 n.9 (1994) (holding that “the defense of ‘superseding cause’ . . . absolves a
tortfeasor, even though his conduct was a substantial contributing factor, when an independent
event intervenes in the chain of causation, producing harm of a kind and degree so far beyond the
risk the original tortfeasor should have foreseen that the law deems it unfair to hold him
responsible”); Perez v. VAS S.p.A., 188 Cal. App. 4th 658, 685 (2010) (noting that “product

5

It is worth noting that neither the plaintiffs nor Apple define the appropriate level of generality
for the duty inquiry. One can imagine other ways of defining the category of defendants and their
relevant conduct: maker of a Bluetooth location-tracking device, maker of a small and affordable
location-tracking device, maker of a device that also tracks location. And the same is true of the
level of generality of the injury. Understandably, courts applying the Rowland factors also often
slip between the general and the more-case-specific, as the Modisette court did. All of which
demonstrates that this duty inquiry is often imprecise. Some slippage is inevitable. But the force
of the conclusions mostly do not hinge on particular nuances of the appropriate level of
generality. And to the extent that they do, it is the party seeking the exception to the baseline rule
of a duty to avoid harm that bears the burden of defining the exception that it seeks. Apple did
not offer a categorical exception—it merely argued that it did not have a duty.

17

misuse may serve as a complete defense when the misuse was so unforeseeable that it should be
deemed the sole or superseding cause” (internal quotations omitted)).
The proximate cause inquiry bears a strong resemblance to the duty inquiry: courts are
meant to consider the connection between the conduct and the injury, as well as any public
policy reasons to limit liability. Modisette, 30 Cal. App. 5th at 153. At the general level, both of
those considerations have already been thoroughly discussed. But the proximate cause inquiry
involves a more specific drill-down: Have the five plaintiffs who were stalked in California
plausibly alleged that the product defects caused their injuries? See Social Media Cases, 2023
WL 6847378, at *29. Unlike duty, proximate cause is typically “a question of fact which cannot
be decided as a matter of law from the allegations of a complaint” unless “the facts are such that
the only reasonable conclusion is an absence of causation.” Modisette, 30 Cal. App. 5th at 152
(quoting State Department of State Hospitals v. Superior Court, 61 Cal.4th 393, 353 (2015)).
Although the complaint certainly could have done a better job of tying each plaintiff’s story to
the alleged defects, the allegations of causation for four of the five California plaintiffs are
sufficient to survive a motion to dismiss.
John Kirkman’s allegations involve several of the purported defects. He alleges that,
despite trying to hide the location of his new home from his estranged wife, she managed to find
the house and to terrorize him and his daughter. The complaint raises an inference that she was
able to locate the home through the use of the AirTag, but it was not until the day after the
incident that Kirkman first received a device notification. Additionally, although both Kirkman
and his daughter allege that they received device notifications, neither of them understood what
the notifications meant, suggesting issues with the clarity of the notification text. Finally, it
allegedly took several days for Kirkman to find the AirTag and remove it—had the sounds been
louder or more distinctive, or had Kirkman had the ability to trigger the sound himself, he could
have found it earlier. Those additional days where Kirkman was searching amounted to
additional time being stalked, during which his estranged wife was allegedly able to see
everywhere that he took his car and identify places that he might go in the future.

18

In Àine O’Neill’s case, the allegations make it clear that the screen notification did not
arrive until after the stalker had discovered a great deal about her: “where she worked, where she
parked her car, her home address, and everywhere she had been.” Several asserted defects with
the product—inadequate noise volume and insufficiently precise tracking—prevented O’Neill
from ever finding the product to trace it back to her stalker. Another asserted defect—inability to
disable the device remotely—prevented O’Neill from being free of the stalking while she still
had her car.6
Hollye Humphreys’s ex-husband allegedly hid an AirTag in the backseat of her car. She
alleges that she received device notifications for three days, but she ignored them because “she
was unfamiliar with AirTags and did not know what the alert meant.” On the third day,
Humphreys received the notification while she was “off work,” and “this time” she “was able to
activate the beeping sound,” which she used to locate the device. And it was only through her
online research that she discovered how to reveal the last four digits of the AirTag owner’s
phone number and was able to confirm that it was her ex-husband. A number of the asserted
defects are implicated in Humphreys’s story: the sound alerts did not begin on their own, and the
sporadic and unclear notifications prolonged the stalking.
However, for one California plaintiff, it is not sufficiently clear from the complaint that
Apple’s alleged breach was the cause of his injuries. Roger Derick Hembd alleges that his wife
repeatedly hid AirTags in his son’s belongings. He describes one incident in March of 2022, in
which he received the AirTag notification during a visitation with his son. He then contacted his
Apple argues that the Court “lacks jurisdiction” over the claims from O’Neill because she is
now a resident of Ireland, even though she was injured while residing in California. The two
cases that Apple cites are both not about jurisdiction and not applicable. The first, Edgar v. MITE
Corp., is a U.S. Supreme Court case that held a state law invalid for burdening interstate
commerce based, in part, on the conclusion that protecting nonresident shareholders is not a
legitimate state interest. 457 U.S. 624, 644 (1982). The second, Ministry of Health, Province of
Ontario, Canada v. Shiley Inc., is a California federal district court case dismissing a products
liability action brought by Canadian plaintiffs against a California manufacturer for forum non
conveniens. 858 F. Supp. 1426 (C.D. Cal. 1994). Notably, Shiley found that the court did have
jurisdiction over the claims, it just concluded that Canada was the appropriate forum for the
claims to be adjudicated. See id. at 1431. Thus, there is no apparent jurisdictional issue with
O’Neill’s claims.
6

19

attorney, who contacted his wife’s attorney, who apologized on behalf of his wife. Id. Months
later, in September of 2022, during another visitation with his son, he allegedly heard a “faint
beeping noise.” After searching for the source of the noise, he found an “altered seam” in his
son’s Batman doll. He tore the seam and found an AirTag—which he discovered was associated
with the last four digits of his wife’s phone number. He filed a police report, the district attorney
brought charges, and a criminal matter was opened; he was also accepted into a confidential
address program administered by the California Secretary of State’s office. Hembd alleges
severe emotional distress following the September 2022 incident and as a result of the AirTag
stalking. But the complaint does not sufficiently connect any of the asserted safety defects—
Apple’s alleged breach—to Hembd’s injury. The plaintiffs do not appear to be arguing that
selling a tracking device is always a violation of the duty of care. And based on the allegations,
the safety mechanisms seem (roughly) to have worked in Hembd’s case. The plaintiffs have not
done enough to draw the connection in the complaint, and thus, they have not adequately alleged
that Apple was the proximate cause of Hembd’s injuries.
For another California plaintiff, the allegations about the role that the AirTag played in
the stalking are too vague. Pamyla Luan was asked by her friend to help him move out of his
home after a breakup and find somewhere else to stay for the short term. Luan gave her friend,
and his belongings, a ride to a hotel. “While she was meeting with her friend, she received a
notification on her iPhone that an unknown AirTag was traveling with her.” But Luan alleges
that neither she nor her friend “understood the significance of the alert.” It allegedly took another
two and a half months before Luan and her friend “found 2 AirTags hidden among her friend’s
belongings”—the same belongings that Luan had put in her car. But removing those AirTags did
not end Luan’s ordeal. She alleges that she was stalked for a year by her friend’s ex-girlfriend,
who repeatedly “sent threatening messages” that “indicated that she was still tracking Luan and
her friend.” The stalker then allegedly took extreme action that further harmed Luan (for
example, illegally obtaining compromising photographs and sending them to Luan’s family, and
sending strangers to Luan’s home to harass her). It is clear that Luan alleges serious injuries as a

20

result of the actions of the woman who stalked her. But it is not clear how the AirTags—or the
asserted safety defects—fit into that narrative. Thus, the plaintiffs have not done enough to show
that Apple’s alleged breach was the proximate cause of Luan’s injuries.
Apple’s own arguments about proximate cause echo its arguments about duty. It relies on
Modisette, which found that no reasonable jury could conclude from the complaint that Apple
caused the injuries to the Modisettes that resulted from a car accident in which the driver was
distracted by FaceTime. Modisette, 30 Cal. App. 5th at 154. The Modisette court, in turn, relied
on multiple cases in which other courts had found that a car accident was not proximately caused
by a product that distracted the driver. Id. at 154–55. But as discussed above, there are
meaningful differences between the iPhone to car accident connection and the AirTag to stalking
connection. Those differences are no less prevalent at the case-specific level. And the car
accident precedent that compelled the Modisette court is less applicable to these facts.
Ultimately, “to the extent that [Apple] can be understood to make the argument that it is
unclear from these factual allegations to what extent [the AirTag and its defects] caused the
specific harm suffered by one of the individual plaintiffs, [Apple] raise[s] a factual argument that
must be addressed at a later stage of the litigation.” Social Media Cases, 2023 WL 6847378, at
*30. That will be a significant question, and a heavily fact-bound one. It is also an inquiry that
seems to vary substantially among plaintiffs. This opinion addresses only five of the thirty-eight
plaintiffs who were joined in this action, and it analyzes those claims under only one state’s tort
law regime—even still, meaningful plaintiff-specific nuances arise in the proximate cause
discussion. These issues do not warrant dismissal at the pleadings stage for the plaintiffs who
have each stated a claim.
III
The plaintiffs also assert claims for strict products liability. They allege that Apple put a
product with safety defects on the market, and that its safety defects caused the plaintiffs’
injuries. Apple argues that these strict products liability claims fail as a matter of law. First, it
asserts that strict products liability cannot apply to these claims due to the nature of the plaintiffs’

21

alleged injuries. Second, it contends that under both tests used to determine whether a product is
“defective” in strict products liability, the plaintiffs’ claims fail as a matter of law.
A
In strict products liability, “a manufacturer is strictly liable in tort when an article he
places on the market . . . proves to have a defect that causes injury to a human being.” Opperman
v. Path, Inc., 87 F. Supp. 3d 1018, 1054 (N.D. Cal. 2014) (quoting Anderson v. Owens-Corning
Fiberglas Corp., 53 Cal.3d 987, 994 (1991)). That rule supplies the duty and breach elements of
a tort claim, leaving only causation and damages. Since the plaintiffs’ negligence claim is largely
based on alleged product defects, the causation analysis comes out the same way for both claims.
See Modisette, 30 Cal. App. 5th at 142. Thus, the key question is whether it is appropriate to
apply strict products liability to these allegations.
Apple argues that strict products liability applies only to allegations of “physical harm to
person or property,” and thus does not cover the emotional distress and psychological harm
alleged by the plaintiffs. In support of this assertion, Apple cites cases describing California’s
economic loss rule. Jimenez v. Superior Court, 29 Cal.4th 473, 481–84 (2002); Opperman, 87 F.
Supp. 3d at 1054–55. The purpose of the economic loss rule is to maintain the “distinction that
the law has drawn between tort recovery for physical injuries and warranty recovery for
economic loss.” Jimenez, 29 Cal.4th at 456. Tort law makes a manufacturer strictly liable for
personal injuries caused by the product. But the law regarding warranties governs “the level of
performance of [the] products in the consumer’s business.” Id. Thus, damages like “inadequate
value” and “costs of repair and replacement of the defective product or consequent loss of
profits” are insufficient to state a products liability claim. Id. Moreover, vague and nonspecific
allegations of damages—for example, just asserting “personal injury” or “invasions of [ ]
privacy”—will also be inadequate. Opperman, 87 F. Supp. 3d at 1054–55.
Apple’s reliance on the economic loss rule is question-begging. Apple is invoking a
doctrine that draws a distinction between economic harm and physical harm to argue that there
must also be a distinction between emotional harm and physical harm. One does not follow from

22

the other. Tort law often uses the term “physical harm” to distinguish from economic harm, and
it has been doing so for a long time. But that only begs the question of how emotional harm
should be treated. Indeed, emotional harm and physical harm are equally distinct from economic
harm, so the economic loss rule should have nothing to say about whether a plaintiff can recover
for emotional harm in a strict products liability case. If there is some other doctrine that
precludes the vindication of an emotional injury in a products liability case, Apple hasn’t cited it.
And so the Court cannot dismiss the plaintiffs’ strict product liability claims on this basis.7
B
California recognizes two different tests for determining whether a product is defective:
the “consumer expectations test” and the “risk-benefit test.” The consumer expectations test asks
whether the products performance is so unsafe “that the defect is apparent to the common reason,
experience, and understanding of its ordinary consumers.” Johnson v. United States Steel Corp.,
240 Cal. App. 4th 22, 32 (2015) (quoting Soule, 8 Cal.4th at 569). That test is “rooted in theories
of warranty,” as “implicit in a product’s presence on the market is a representation that it is fit to
do safely the job for which it was intended.” McCabe v. American Honda Motor Co., 100 Cal.
App. 4th 1111, 1120 (2002). This test is a poor fit for the plaintiffs’ claims. The asserted issue
with the product is not that it is unsafe for the purchasers—the issue is that it is unsafe for people
the purchasers hurt by using the product to perform its one function (location tracking).
Moreover, the consumer expectations test is inappropriate for certain complex products if
“the circumstances of the product’s failure” do not themselves “permit an inference that the
product’s design performed below” the reasonable safety expectations of consumers. Marmont v.
Bernzomatic Corp., No. 16-cv-848-JAK, 2018 WL 6252500, at *16 (C.D. Cal. July 31, 2018)
7

There is also cause for skepticism that California courts would categorically bar product
liability recovery for pure emotional distress injury. Imagine a manufacturer sells a virtual reality
headset designed for meditation. And imagine that a product defect causes the headset to glitch,
resulting in the display of violent and disturbing images that cause serious anxiety for the user. If
all the other elements for a products liability claim were present, it is difficult to believe that tort
recovery would be barred. But even if that were the appropriate result under California law, that
result is certainly not compelled by the economic loss doctrine—which is all that Apple has
pointed to.

23

(quoting Demara v. Raymond Corp., 13 Cal. App. 5th 545, 559 (2017)). And the circumstances
of the stalking incidents are not enough to demonstrate defects with the AirTag itself. Rather,
issues like the inability of iPhone owners to initiate a scan for AirTags, or the requirement of
physical possession to disable the AirTag, will require an investigation into the technology. This
is a situation in which an ordinary consumer “would have no idea how safe the product could be
made.” McCabe, 100 Cal. App. 4th at 1121 (quoting Soule, 8 Cal.4th at 562). For both of these
reasons, the risk-benefit test is a better fit, and, for these asserted product defects, the consumer
expectations test is unavailable as a matter of law. Cf. Marmont, 2018 WL 6252500, at *17.
Under the risk-benefit test, products “may be defective if the design embodies an
excessive preventable danger.” McCabe, 100 Cal. App. 4th at 1121. In other words, the
factfinder will ask whether the “risks of the design outweigh its benefits.” Id. The plaintiff has
the initial burden of demonstrating that the design was the proximate cause of their injuries.
After that, “the burden shifts to the defendant to establish that the benefits of the challenged
design, when balanced against such factors as the feasibility and cost of alternative designs,
outweigh its inherent risk of harm.” Id. Apple argues that the plaintiffs have not stated a strict
products liability claim that can be adjudicated under the risk-benefit test because “they do not
plead the existence of a feasible alternative design.” But the burden is not on the plaintiffs to
identify what cost-justified measures Apple could have taken to cure the defects. That makes
sense—that information is typically in the hands of the manufacturer, not the party injured by the
manufacturer’s product. Additionally, the plaintiffs have actually identified a number of things
that they think that Apple could do to make the product safer. Throughout the briefing for the
motion to dismiss, Apple asserts that it has made the product as safe as reasonably possible, that
it has been innovating and adding features as problems have occurred, and that the safety
features make it unforeseeable that people are harmed by AirTag stalking. Under the risk-benefit
test, Apple will have a chance to demonstrate all of that.
IV
The motion to dismiss is denied with regard to the negligence claims and risk-benefit

24

products liability claims asserted by Kirkman, O’Neill, and Humphreys. The same claims
brought by Hembd and Luan are dismissed with leave to amend. As noted in the accompanying
order, the rest of the plaintiffs’ claims are also dismissed with leave to amend. Discovery may
move forward immediately on the surviving claims. Any amended complaint is due within 21
days of this ruling.
IT IS SO ORDERED.
Dated: March 15, 2024
______________________________________
VINCE CHHABRIA
United States District Judge

25