`
`Mohammad Tajsar (SBN 280152)
`mtajsar@aclusocal.org
`ACLU Foundation of Southern California
`1313 West 8th Street
`Los Angeles, CA 90017
`Telephone: (213) 977-9500
`Facsimile: (213) 977-5297
`
`Jacob A. Snow (SBN 270988)
`jsnow@aclunc.org
`ACLU Foundation of Northern California
`39 Drumm Street
`San Francisco, CA 94111
`Telephone: (415) 621-2493
`Facsimile: (415) 255-8437
`
`Counsel for Plaintiffs
`(continued on next page)
`
`
`
`
`
`
`
`
`
`UNITED STATES DISTRICT COURT FOR THE
`
`CENTRAL DISTRICT OF CALIFORNIA
`
`CASE NO: 2:20-cv-05044
`
`COMPLAINT
`
`
`
`
`
`Plaintiffs,
`
`JUSTIN SANCHEZ and ERIC ALEJO;
`
`
`v.
`
`LOS ANGELES DEPARTMENT OF
`TRANSPORTATION and CITY OF
`LOS ANGELES,
`
` Defendants.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`
`
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 2 of 19 Page ID #:2
`
`(continued from previous page)
`
`
`Jennifer Lynch (SBN 240701)
`jlynch@eff.org
`Electronic Frontier Foundation
`815 Eddy Street
`San Francisco, CA 94109
`Tel: (415) 463-9333
`Fax: (415) 436-9993
`
`Douglas E. Mirell (SBN 94169)
`DMirell@ggfirm.com
`Timothy J. Toohey (SBN 140117)
`TToohey@ggfirm.com
`Greenberg Glusker Fields Claman & Machtinger LLP
`2049 Century Park East, Suite 2600
`Los Angeles, California 90067
`Telephone: (310) 553-3610
`Fax:
`(310) 553-0687
`
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`
`
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 3 of 19 Page ID #:3
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`
`
`INTRODUCTION
`
`1.
`
`Beginning in late 2017, communities across California witnessed a
`
`near-overnight invasion of motorized electric scooters on city sidewalks. Equipped
`
`with tiny motors, batteries, and the sleek insignia of their proprietor technology
`
`companies, they introduced a new dockless mode of transit for smartphone-
`
`equipped consumers as an alternative to cars, bicycles, and public transit. Similar
`
`to a car ride-share service, riders reserve and pay for scooter rentals through a
`
`smartphone app. At the end of a trip, the user leaves the scooter on the street,
`
`where it can be rented again.
`
`10
`
`2.
`
`Soon after scooters appeared, complaints targeting the scooter
`
`11
`
`companies followed. Although dockless scooters represented a novel and
`
`12
`
`potentially useful form of transit, they also cluttered city sidewalks, lacked safety
`
`13
`
`features, and interfered with disabled access to city streets. The scooter companies
`
`14
`
`themselves often did jurisdictions no favors, aggressively pushing back against
`
`15
`
`attempts to regulate the vehicles.
`
`16
`
`3.
`
`As in other cities across the country, this was the story of scooters in
`
`17
`
`Los Angeles. In an attempt to avoid the unpopular profusion of scooters filling the
`
`18
`
`sidewalks, Defendants Los Angeles Department of Transportation and the City of
`
`19
`
`Los Angeles (collectively “LADOT” or “Defendants”) developed a far-reaching
`
`20
`
`software tool that (they claim) is necessary to managing the right of way. Dubbed
`
`21
`
`the Mobility Data Specification (“MDS”), this software interface, crafted in
`
`22
`
`partnership with a private consultancy, forces operators of dockless vehicles to
`
`23
`
`provide real-time and historical data about each vehicle and trip taken in Los
`
`24
`
`Angeles, all as a condition of operating. Most importantly, the tool requires that
`
`25
`
`scooter companies produce detailed trip data about every single scooter trip taken
`
`26
`
`within city limits, including where each trip starts, the route it takes, and where it
`
`27
`
`ends.
`
`4.
`
`Although MDS does not record the identity of the rider directly, the
`
`28
`
`
`
`COMPLAINT
`1
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 4 of 19 Page ID #:4
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`precision with which it captures riders’ location information—often to within a few
`
`
`feet—likely allows riders to be identified. Knowing that a particular trip began at
`
`an office building and ended in front of a home, for example, makes the difficulty
`
`of identifying the individual rider as simple as knowing their home and work
`
`addresses. Given the large amount of public or otherwise accessible data about
`
`people’s lives that exists, simply cross-referencing MDS data about a particular
`
`trip with any other dataset (including mere observation of a routinely-taken scooter
`
`trip) can reveal who took the trip.
`
`5.
`
`Beyond identifying an individual rider, the locations where an
`
`10
`
`individual’s trip starts and ends can also reveal why that rider made the trip.
`
`11
`
`Regular trips that start near a residence and end at an office reveal that a person
`
`12
`
`living at the residence works at the office and takes a particular route to work.
`
`13
`
`Periodic trips that begin at a high school and end in a family-planning clinic could
`
`14
`
`reveal that a student is seeking reproductive health care. Even a single trip to a
`
`15
`
`protest against police violence may result in a rider’s name being revealed and her
`
`16
`
`presence at the protest exposed against her wishes.
`
`17
`
`6.
`
`LADOT has never articulated an adequate or reasonable justification
`
`18
`
`for the collection of such sensitive location information en masse. When mandated
`
`19
`
`by the Los Angeles City Council to identify, by February 25, 2020, its reasons for
`
`20
`
`collecting precise location data, LADOT failed to do so. Now, over three months
`
`21
`
`after this deadline, LADOT has still not articulated an operationally specific need
`
`22
`
`for this data. To date, it has offered only the most generic justifications for
`
`23
`
`collecting precise location information, stating at one point that its goal is to
`
`24
`
`“experiment” with riders’ protected information when setting agency policy.
`
`25
`
`7.
`
`The Constitution prohibits LADOT from experimenting with the
`
`26
`
`rights of its constituents. The Fourth Amendment strictly limits the warrantless
`
`27
`
`collection of vehicular location information. As a Supreme Court majority
`
`28
`
`recognized in United States v. Jones, “GPS monitoring generates a precise,
`
`
`
`COMPLAINT
`2
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 5 of 19 Page ID #:5
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`comprehensive record of a person’s public movements that reflects a wealth of
`
`
`detail about her familial, political, professional, religious, and sexual associations.”
`
`565 U.S. 400, 415 (2012) (J., Sotomayor, concurring); id. at 430 (J., Alito,
`
`concurring) (long-term capture of vehicle location information violates reasonable
`
`expectation of privacy). This is particularly true here, where the scale and breadth
`
`of that data collection has no conceivable relation to a targeted investigation of a
`
`particular individual. MDS collects precise location data associated with every
`
`single rider of scooters within the City, every single time they ride such a vehicle.
`
`And once MDS software is deployed, it gathers location data without any human
`
`10
`
`involvement and at the maximum precision generated by the vehicles.
`
`11
`
`8.
`
`Plaintiffs ride electric scooters in the City of Los Angeles, using the
`
`12
`
`vehicles to make trips from their homes to work, friends, businesses, and places of
`
`13
`
`leisure. LADOT uses MDS to warrantlessly collect sensitive vehicle location data
`
`14
`
`associated with each of Plaintiffs’ trips, in violation of their right to be free from
`
`15
`
`unreasonable searches and seizures in contravention of the United States and
`
`16
`
`California Constitutions. The compelled production of Plaintiffs’ location
`
`17
`
`information also violates the California Electronic Communications Privacy Act
`
`18
`
`(“CalECPA”).
`
`19
`
`9.
`
`LADOT violates these rights irrespective of whether it collects data
`
`20
`
`about Plaintiffs’ movements in real-time or after a period of delay. The gathering
`
`21
`
`of historical location information about individuals without sufficient justification
`
`22
`
`violates the Constitution. United States v. Carpenter, 138 S. Ct. 2206, 2218 (2018)
`
`23
`
`(comparing the greater harms of historical location tracking as opposed to manual
`
`24
`
`real-time observation, and explaining that “[u]nlike with the GPS device in Jones,
`
`25
`
`police need not even know in advance whether they want to follow a particular
`
`26
`
`individual, or when.”). When that location data is highly precise (as the MDS data
`
`27
`
`is), the risks with collecting historical location information are too great without a
`
`28
`
`warrant.
`
`
`
`COMPLAINT
`3
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 6 of 19 Page ID #:6
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`
`
`10. Through this action, Plaintiffs seek, among other forms of relief,
`
`expungement of their location data from LADOT’s servers and an injunction
`
`preventing MDS from collecting and storing precise location information en
`
`masse.
`
`JURISDICTION
`
`11. This Court has subject matter jurisdiction over this action under 28
`
`U.S.C. § 1331 and 28 U.S.C. § 1343 because it alleges violations of the United
`
`States Constitution enforceable through 42 U.S.C. § 1983, and under 28 U.S.C.
`
`§ 1367 because it alleges violations of the California Constitution and the
`
`10
`
`California Penal Code.
`
`11
`
`12
`
`12. Venue is proper under 28 U.S.C. § 1391(a) because Defendants are
`
`VENUE
`
`13
`
`residents of this district, 28 U.S.C. § 1391(b) because the events and actions giving
`
`14
`
`rise to the claims herein occurred in this District, and 28 U.S.C. § 1391(c) because
`
`15
`
`Defendants are subject to the court’s personal jurisdiction in this District.
`
`16
`
`17
`
`13. Plaintiff Justin Sanchez is a resident of Los Angeles, and a customer
`
`PARTIES
`
`18
`
`of and rider of dockless vehicles offered by Lime, Bird, and Lyft. Mr. Sanchez has
`
`19
`
`ridden scooters operated by these three providers within the City of Los Angeles
`
`20
`
`while MDS has been in effect. Mr. Sanchez intends to continue riding these
`
`21
`
`dockless vehicles within Los Angeles in the future.
`
`22
`
`14. Plaintiff Eric Alejo is a resident of Los Angeles, and a customer of
`
`23
`
`and rider of dockless vehicles offered by Lyft, JUMP, Bird, and Lime. Mr. Alejo
`
`24
`
`has ridden scooters operated by these four providers within the City of Los
`
`25
`
`Angeles while MDS has been in effect. Mr. Alejo also intends to continue riding
`
`26
`
`these dockless vehicles within Los Angeles in the future.
`
`27
`
`15. Defendant City of Los Angeles is a public entity, duly organized and
`
`28
`
`existing under the laws of the State of California.
`
`
`
`COMPLAINT
`4
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 7 of 19 Page ID #:7
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`
`
`16. Defendant Los Angeles Department of Transportation is a government
`
`agency created by Los Angeles city ordinance. It advertises that its mission is “to
`
`lead transportation planning, project delivery, and operations in the City of Los
`
`Angeles.” LADOT developed and operates the permitting program that licenses the
`
`operation of dockless vehicle providers within Los Angeles.
`
`STATEMENT OF FACTS
`
`I.
`
`DOCKLESS VEHICLES LAUNCH IN SOUTHERN CALIFORNIA IN
`
`2017, AND ARE FORMALLY ALLOWED IN THE CITY OF LOS
`
`ANGELES BEGINNING IN MARCH 2019.
`
`10
`
`17. Beginning in 2017, numerous private companies began dropping
`
`11
`
`dockless vehicles, including motorized scooters and electric bicycles, on Los
`
`12
`
`Angeles streets. These vehicles are owned and maintained by private companies,
`
`13
`
`and individual customers can rent them via a smartphone application. These
`
`14
`
`vehicles are “dockless” in that rides need not start from a fixed docking station,
`
`15
`
`like a traditional municipal bicycle share. Instead, rides can begin and end
`
`16
`
`wherever a vehicle is located, with a user employing only the provider’s mobile
`
`17
`
`application to terminate the rental. At that point, the application informs the rider
`
`18
`
`of the cost of the ride, and charges the user accordingly.
`
`19
`
`18. The operators typically outfit the vehicles with rechargeable batteries,
`
`20
`
`Global Positioning System (“GPS”) trackers, and wireless connectivity to the
`
`21
`
`internet. The vehicles broadcast precise GPS coordinates to the operator, which
`
`22
`
`allows it to track rides and charge customers accordingly.
`
`23
`
`19. With the growing use of dockless vehicles in neighboring cities, the
`
`24
`
`Los Angeles City Council passed an ordinance on September 28, 2018 compelling
`
`25
`
`LADOT to implement a “Shared Mobility Device Pilot Program” establishing an
`
`26
`
`application process for the approval of City-issued permits to operators of dockless
`
`27
`
`bicycles, electric bicycles, motorized scooters, and electric scooters. The ordinance
`
`28
`
`mandated that under the Pilot Program, “an operator of a shared mobility device
`
`
`
`COMPLAINT
`5
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 8 of 19 Page ID #:8
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`shall obtain a permit from the Department [of Transportation] and comply with all
`
`
`Department permit rules, regulations, indemnification, insurance and fee
`
`requirements.”
`
`20.
`
`In response, LADOT created the permitting program via an
`
`application and review process. In exchange for a license to operate a
`
`micromobility company within City limits, LADOT instructed operators to submit
`
`an application detailing their intended deployment in the City and agreeing to
`
`numerous regulatory requirements.1 These requirements included otherwise
`
`standard insurance requirements, an agreement to an indemnification provision,
`
`10
`
`limits on the total number of vehicles any individual operator could deploy within
`
`11
`
`Los Angeles, agreements to place safety features like lights and reflectors on
`
`12
`
`vehicles, requirements that operators ensure vehicles are appropriately parked and
`
`13
`
`not blocking pedestrian rights of way, and incentives for operators to diversify the
`
`14
`
`geographic distribution of their vehicles. Relevant here, the permitting application
`
`15
`
`required operators agree to implement MDS’s data collection protocols.
`
`16
`
`21. Once an operator applied for a permit, LADOT reviewed the
`
`17
`
`application and awarded an operating permit to the applicant accordingly.
`
`18
`
`Individual end users were not parties to the application agreement, despite their
`
`19
`
`sensitive location data being critical to the process.
`
`20
`
`22. The original one-year pilot program launched in March 2019, and has
`
`21
`
`been extended for six months through September 15, 2020. At the close of the
`
`22
`
`now-eighteen-month pilot, LADOT plans to establish a one-year formal dockless
`
`23
`
`mobility pilot program.
`
`24
`
`II. DEFENDANTS UTILIZE THE MOBILITY DATA SPECIFICATION
`
`TO UNLAWFULLY COLLECT PRECISE MOVEMENT DATA.
`
`23. As a condition of securing a permit to operate in the City of Los
`
`
`1 LADOT’s original permit application is available at
`https://files.acluwest.org/s/XATp4ErkW4WsSsT.
`
`25
`
`26
`
`27
`
`28
`
`
`
`COMPLAINT
`6
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 9 of 19 Page ID #:9
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`Angeles, LADOT requires mobility companies to implement MDS’s data
`
`
`collection requirements. MDS contains two interrelated parts: 1) the data-
`
`collection standard, which specifies what information mobility providers must
`
`deliver to the governing jurisdiction, and 2) reference implementations in software
`
`code that both mobility providers and the governing jurisdiction can use to set up
`
`the information exchange. MDS ingests data directly from the transportation
`
`companies, enabling LADOT “to actively manage private mobility providers and
`
`the public right-of-way . . . through a shared data vocabulary and to communicate
`
`directly with product companies in real time using code.”2
`
`10
`
`24. The purpose of MDS is to accelerate information collection by cities
`
`11
`
`and counties facing an increase in the volume of permitting associated with
`
`12
`
`dockless scooters. According to the non-profit Open Mobility Foundation, the
`
`13
`
`proprietor of MDS who took over its administration from LADOT, “the goals of
`
`14
`
`MDS are to provide a standardized way for municipalities or other regulatory
`
`15
`
`agencies to ingest, compare and analyze data from mobility service providers, and
`
`16
`
`to give municipalities the ability to express regulation in machine-readable
`
`17
`
`formats. . . . MDS is a key piece of digital infrastructure that supports the effective
`
`18
`
`implementation of mobility policies in cities around the world.” Instead of each
`
`19
`
`city deciding for itself what information to collect and writing the necessary
`
`20
`
`software, MDS encourages cities to adopt a single existing standard.
`
`21
`
`25. MDS, once implemented by private dockless scooter companies,
`
`22
`
`ingests a wide variety of data directly from the providers without any human input.
`
`23
`
`The data includes the provider’s name, a unique device identifier for the vehicle,
`
`24
`
`the type of vehicle, the length of the trip, its starting point, end point, and the route
`
`25
`
`the vehicle took on its trip. Relevant here is the route information requirement,
`
`
`2 “Mobility Data Specification: Information Briefing,” Los Angeles
`Department of Transportation, https://ladot.io/wp-content/uploads/2018/12/What-
`is-MDS-Cities.pdf, Oct. 31, 2018.
`
`26
`
`27
`
`28
`
`
`
`COMPLAINT
`7
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 10 of 19 Page ID #:10
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`which calls for granular trip data from the providers to LADOT about every ride
`
`
`taken within Los Angeles—including the starting point of the ride, the starting time
`
`of the ride, the end point for the ride, and the ending time of the ride. LADOT
`
`requires that start and end locations be provided in real-time, and the route that the
`
`trip took between those points provided after 24 hours.
`
`26. While MDS does not collect any information directly identifying the
`
`rider of a particular vehicle, the sensitivity of movement information makes it
`
`possible to identify individual riders anyway. Coupling a rider’s precise trip data
`
`with information from just one other dataset—for instance, additional scooter rides
`
`10
`
`that show a pattern of repeated trips to and from the same locations, public voting
`
`11
`
`records from particular addresses, or even simple physical observation of a rider—
`
`12
`
`can likely identify the individual who took the trip. In addition, it may reveal
`
`13
`
`important information about the individual’s residence, the identity of her
`
`14
`
`employer, associates, or friends, the type of physicians she visits, or her favorite
`
`15
`
`recreational activities. And when end points are sensitive locations—like
`
`16
`
`therapists’ offices, marijuana dispensaries, or Planned Parenthood clinics—those
`
`17
`
`routes may reveal why she made that trip.
`
`18
`
`27.
`
`In a time when protests are erupting around the country, the risk of
`
`19
`
`identifying individuals based on physical observation takes on a new importance.
`
`20
`
`Imagine a person who takes a scooter to a political protest, or even rides past and is
`
`21
`
`captured by one of the many cameras used to document interaction between
`
`22
`
`protesters and police. With the information LADOT ingests through MDS, that
`
`23
`
`individual ride could be picked out of a haystack of data and handed over to the
`
`24
`
`police, who would then know where the person ended their trip, where they started,
`
`25
`
`and the precise route they took.
`
`26
`
`28. The likelihood of identifying individuals based only on location
`
`27
`
`information is not a hypothetical concern. A growing body of research has
`
`28
`
`demonstrated that location datasets are easily susceptible to identification. “With
`
`
`
`COMPLAINT
`8
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 11 of 19 Page ID #:11
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`merged mobility datasets, this becomes even easier: An agent could potentially
`
`
`match users trajectories in anonymized data from one dataset, with deanonymized
`
`data in another, to unmask the anonymized data.”3 For instance, researchers have
`
`found that they could identify 50% of people from only two randomly chosen data
`
`points in a dataset that contained only time and location data.4
`
`29.
`
`Identification of location data poses grave risks to individuals—
`
`particularly marginalized or justice-impacted members of the community. In
`
`addition to revealing sensitive information about people’s lives, this information
`
`can exacerbate persistent forms of state violence and bias that target those at the
`
`10
`
`fringes. This includes police encounters, immigration enforcement, homelessness
`
`11
`
`sweeps, or enforcement of pre-trial release terms or probation conditions, to name
`
`12
`
`a few. Past experience has also shown that individual location information in the
`
`13
`
`hands of authorities can stoke racial and gender-based violence. When collected
`
`14
`
`without adequate safeguards, location information often results in cases of
`
`15
`
`domestic abuse and stalking, as a recent investigation of automatic license plate
`
`16
`
`reader information in California revealed.5
`
`
`3 Rob Matheson, The privacy risks of compiling mobility data: Merging
`different types of location-stamped data can make it easier to discern users’
`identities, even when the data is anonymized, MIT News (Dec. 7, 2018),
`http://news.mit.edu/2018/privacy-risks-mobility-data-1207 (describing Daniel
`Kondor et al., “Towards matching user mobility traces in large-scale datasets,”
`IEEE Transactions on Big Data (Sep. 24, 2018), available at
`http://senseable.mit.edu/papers/pdf/20180927_Kondor-
`etal_TowardsMatching_IEEE-BigData.pdf)).
`4 Yves-Alexandre de Montjoye, et al., Unique in the Crowd: The privacy
`bounds of human mobility, 3 Nature Scientific Reports 1376 (2013),
`http://www.nature.com/articles/srep01376 (finding that “in a dataset where the
`location of an individual is specified hourly, and with a spatial resolution equal to
`that given by the carrier’s antennas, four spatio-temporal points are enough to
`uniquely identify 95% of the individuals.”).
`5 “Automated License Plate Readers: To Better Protect Individuals’’
`Privacy, Law Enforcement Must Increase Its Safeguards for the Data It Collects,”
`(cont’d)
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`COMPLAINT
`9
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 12 of 19 Page ID #:12
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`
`
`30. Despite the sensitivity of location data and the increasing legal
`
`protections individuals have over their location information, LADOT’s MDS
`
`protocol demands maximally precise locations about individuals’ trips. LADOT’s
`
`technical consultants developed the program to collect as precise information as
`
`the vehicles generate. With the pilot program in full effect, LADOT captured GPS
`
`coordinates broadcast by scooters up to seven decimal places, an extraordinary
`
`level of accuracy even assuming a wide margin of error. For background, GPS
`
`coordinates are often expressed through decimal degrees via longitude and latitude
`
`coordinates. The more decimal places a GPS coordinate is measured in, the more
`
`10
`
`precise the location it reveals is.6 MDS compels vehicle operators to provide the
`
`11
`
`coordinates of each vehicle in latitude and longitude to the maximum precision
`
`12
`
`allowed by the vehicle, which can be up to seven decimal places. For reference,
`
`13
`
`coordinates in seven decimal degrees are accurate to within 1.11 centimeters at the
`
`14
`
`equator. In real-world conditions away from the equator, the accuracy with which
`
`15
`
`the companies capture a vehicle’s location depends on the scooter’s hardware, the
`
`16
`
`availability of over-the-air internet connectivity, and the physical conditions
`
`17
`
`surrounding a vehicle that may impact GPS signal strength. Given advances in
`
`18
`
`technology and widespread access to 4G (and, soon, 5G connectivity), the
`
`19
`
`coordinates generated by dockless vehicles can accurately place them within a few
`
`20
`
`dozen feet, indicating with confidence, for example, where on a city block a
`
`21
`
`scooter is and the building or piece of city infrastructure nearest which it is parked.
`
`22
`
`31. Research has demonstrated that even truncating GPS coordinates of
`
`23
`
`trips’ origins and destinations by lopping off GPS decimal places does little to
`
`
`California State Auditor (Feb. 20, 2020),
`https://www.auditor.ca.gov/pdfs/reports/2019-118.pdf, at 12–13 (discussing
`instance of gender-based assault resulting from license plate location information).
`6 For a helpful explanation of the math behind the precision of GPS
`locations, see “Decimal Degrees,” WIKIPEDIA,
`https://en.wikipedia.org/wiki/Decimal_degrees (last visited June 3, 2020).
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`COMPLAINT
`10
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 13 of 19 Page ID #:13
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`protect individuals’ privacy. Such is the sensitivity of location data sets, which
`
`
`cannot reasonably be considered “anonymized” in any real sense when collected en
`
`masse and with the precision that MDS currently demands.
`
`32. Given these facts, and upon information and belief, a simple analysis
`
`of MDS data will likely identify the precise trips taken by Plaintiffs in this case and
`
`where they live, work, shop, and frequent. Plaintiffs have never agreed to share
`
`their precise location data with LADOT, even though LADOT has used MDS to
`
`extract this data from the operators whose vehicles Plaintiffs rented.
`
`III. DEFENDANTS HAVE FAILED TO PROVIDE REASONABLE
`
`JUSTIFICATIONS FOR COMPELLING PRODUCTION OF
`
`LOCATION INFORMATION.
`
`33. When LADOT launched its pilot MDS program, it did not identify
`
`13
`
`with operational specificity how it intended to use the data ingested by MDS. Nor
`
`14
`
`did it develop a data collection and retention program narrowly tailored to meet
`
`15
`
`even the use cases it did identify. As a result, granular location information easily
`
`16
`
`susceptible to identification is needlessly—and illegally—collected and at risk of
`
`17
`
`being shared with third parties and targeted by other government actors.
`
`18
`
`34. According to LADOT, the purpose of MDS is multifaceted, and
`
`19
`
`allows LADOT to “actively manage private companies who operate in our public
`
`20
`
`space.”7 During the development of the pilot program, LADOT identified
`
`21
`
`numerous overlapping and related benefits that mass location data may provide to
`
`22
`
`the regulators, some more specific than others, but none that necessitated collecting
`
`23
`
`all riders’ granular and precise location information en masse.
`
`35. To the contrary, LADOT leadership expressly identified the MDS
`
`pilot as a mechanism to “experiment” with this data collection project.8 Put
`
`7 “LADOT | Putting Ideas into motion,” https://ladot.io/ (embedding
`YouTube video entitled “The Future of Mobility – Mobility Data Specification”).
`8 David Zipper, “Cities can see where you’re taking that scooter,” Slate, Apr.
`(cont’d)
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`COMPLAINT
`11
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 14 of 19 Page ID #:14
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`differently, LADOT has intentionally gathered sensitive and legally protected
`
`
`location information in an experimental posture with data concerning actual users
`
`in real time—unlawfully collecting data in order to determine post hoc how to
`
`exploit them to serve non-specific regulatory needs.
`
`36. LADOT’s dockless scooter pilot program fails to tailor its collection
`
`of individual location data to articulable use cases. Given the lack of specificity for
`
`its proposed use cases, it is difficult to identify potential alternatives to maintaining
`
`individual trip data, or to gauge the necessity of mass location data collection to
`
`achieve LADOT’s regulatory ends.
`
`10
`
`37. To the contrary, LADOT has been exceptionally vague about how it
`
`11
`
`intends to use this data, in part to allow the agency to experiment with exploiting
`
`12
`
`the data after it is collected. Each of the articulated use cases LADOT has offered
`
`13
`
`for its desire to collect en masse individual vehicle location data fails under
`
`14
`
`scrutiny. For instance, the City Council mandated that LADOT’s pilot program
`
`15
`
`incentivize providers to diversify access to its vehicles. Addressing equity in
`
`16
`
`regional distributions of vehicles by itself does not require individual, granular
`
`17
`
`location data. For instance, collecting a vehicle’s neighborhood-level locations at
`
`18
`
`regular, but disparate, time intervals (e.g., every two hours) will adequately inform
`
`19
`
`regulators whether providers are distributing their vehicles equitably—without
`
`20
`
`collecting individuals’ trip data.
`
`21
`
`38. Even the few use cases LADOT has offered for why granular trip data
`
`22
`
`is necessary do not require detailed vehicle telemetry data. For instance, even
`
`23
`
`though MDS calls for acquiring GPS coordinates up to a maximum level of
`
`24
`
`precision, current physical limitations on the accuracy of GPS broadcasts from
`
`25
`
`26
`
`27
`
`28
`
`
`
`
`2, 2019, https://slate.com/business/2019/04/scooter-data-cities-mds-uber-lyft-los-
`angeles.html (quoting LADOT General Manager Seleta Reynolds, “When bikes
`and scooters showed up, they gave us a pretty interesting sandbox to start
`experimenting.”)
`
`COMPLAINT
`12
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 15 of 19 Page ID #:15
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`vehicles make their coordinates too imprecise to determine whether scooters are
`
`
`appropriately parked adjacent to a curb versus inappropriately parked in the middle
`
`of a sidewalk a couple of feet away, another purpose LADOT has offered for why
`
`it needs individual users’ trip information.
`
`39. Even when the City Council instructed LADOT to articulate “specific
`
`regulatory purposes for the collection and use of each type of data required by
`
`MDS,” LADOT simply did not comply. This request came in a Los Angeles City
`
`Council Motion passed on November 27, 2019, which required LADOT to, among
`
`other things, provide a report outlining the “specific regulatory purposes for the
`
`10
`
`collection and use of each type of data required by MDS” by February 25, 2020.9
`
`11
`
`To date, and more than three months after the deadline, LADOT has still not
`
`12
`
`articulated those purposes in response to the request.
`
`13
`
`40.
`
`Importantly, LADOT lacks reasonable and justifiable uses for both
`
`14
`
`real-time location information and historical location data—both of which present
`
`15
`
`substantially similar violations of Plaintiffs’ rights. This is in contrast to concerns
`
`16
`
`raised by some of the micromobility providers, including JUMP. While JUMP has
`
`17
`
`historically protested LADOT’s requirement that it produce precise location
`
`18
`
`information in real-time, including by filings its own lawsuit challenging MDS
`
`19
`
`data collection by MDS, JUMP offered to produce that information to LADOT
`
`20
`
`after a 24-hour delay.10 According to JUMP, a 24-hour delay “significantly
`
`21
`
`mitigate[s] the frightening risks of direct and constant government surveillance and
`
`22
`
`possible interception of individual users.”11
`
`41. Perhaps most ominously, LADOT plans to extend the same model for
`
`
`9 Motion No. 19-1355, Intro. By David Ryu (Nov. 1, 2019), available at
`http://clkrep.lacity.org/onlinedocs/2019/19-1355_mot_11-01-2019.pdf.
`10 Social Bicycles LLC d/b/a JUMP v. City of Los Angeles, No. 2:20-cv-
`02746 (C.D. Cal. filed Mar. 24, 2020), ECF No. 1, at ¶ 11 (challenging MDS
`location collection from perspective of private scooter operator).
`11 Id.
`
`23
`
`24
`
`25
`
`26
`
`27
`
`28
`
`
`
`COMPLAINT
`13
`
`
`
`
`Case 2:20-cv-05044-CJC-PD Document 1 Filed 06/08/20 Page 16 of 19 Page ID #:16
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`real-time geolocation information collection that it uses for dockless vehicles to
`
`
`“all kinds of future transportation forms—