`
`
`
`
`
`
`
`
`
`
`
`
`
`IN THE SUPERIOR COURT OF WASHINGTON FOR KING COUNTY
`
`
`
`MICHELE ROSATI,
`
`Plaintiff,
`
`
`
`
`NO.
`
`COMPLAINT FOR RELIEF PURSUANT
`TO 8 DELAWARE GENERAL
`CORPORATION LAW CODE SECTION
`220 TO COMPEL INSPECTION OF
`BOOKS AND RECORDS
`
`v.
`
`AMAZON.COM, INC., a Delaware
`corporation,
`
`
`Defendant.
`
`
`
`
`Plaintiff Michele Rosati ("Plaintiff") herein alleges, upon knowledge as to herself and
`
`her own actions, and upon information and belief as to all other matters, as follows:
`
`I.
`
`NATURE OF THE ACTION
`
`1.1
`
`In this action, Plaintiff seeks to enforce her right to inspect certain corporate
`
`books and records of defendant Amazon.com, Inc. ("Amazon" or the "Company"), a Delaware
`
`corporation, pursuant to title 8, section 220 of the Delaware General Corporation Law Code
`
`("Section 220"). Plaintiff is a beneficial stockholder of the Company.
`
`1.2
`
`On September 22, 2020, Plaintiff sent an inspection demand to the Company (the
`
`"Inspection Demand"). The Inspection Demand complied with all the form and manner
`
`requirements of Section 220, including that it was accompanied by a power of attorney, an oath,
`
`COMPLAINT - 1
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`FILED
`
`2021 MAR 18 02:26 PM
`
`KING COUNTY
`
`SUPERIOR COURT CLERK
`
`E-FILED
`
`CASE #: 21-2-03591-1 SEA
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 2 of 46
`
`
`
`and proof of Plaintiff's current ownership of Amazon stock. The Company received the letter
`
`on September 23, 2020.1
`
`1.3
`
`In the Inspection Demand, Plaintiff explained that she wishes to investigate
`
`potential wrongdoing occurring at the Company, including potential breaches of fiduciary duty.
`
`Plaintiff has legitimate concerns as to whether Amazon violated the Illinois Biometric
`
`Information Privacy Act ("BIPA") and whether the Company was engaging in antitrust
`
`violations.
`
`1.4
`
`The Inspection Demand provided ample evidence of such possible wrongdoing
`
`and mismanagement at Amazon. Concerning BIPA, the Inspection Demand explained that
`
`Amazon had been developing facial recognition software for years and purchased International
`
`Business Machines Corporation's ("IBM") "Diversity in Faces" dataset in 2019 to improve this
`
`software. In developing this facial recognition software, Amazon collected, stored, and used
`
`individuals' biometric identifiers without ever informing those before, a direct violation of
`
`BIPA.
`
`1.5
`
`Regarding Amazon's anticompetitive violations, the Inspection Demand again
`
`contained detailed information how the Company uses third-party seller data it has access to as
`
`an effective middleman to develop its own competing suite of products. The Company then
`
`undercuts the third-party on price. Amazon's anticompetitive actions have led to investigations
`
`by, at a minimum: (i) the U.S. Congress; (ii) the European Union; (iii) the State of California;
`
`and (iv) the State of Washington. Accordingly, Plaintiff has ample reason to suspect
`
`wrongdoing at Amazon, more than satisfying the credible basis standard necessary to justify the
`
`inspection.
`
`
`1 True and correct copies of the Inspection Demand and proof of delivery are attached hereto as Exhibit A and B, respectively.
`
`COMPLAINT - 2
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 3 of 46
`
`
`
`1.6
`
`On October 19, 2020, the Company's counsel responded to Plaintiff's Inspection
`
`Demand with a short perfunctory two-and-a-half-page rejection. The rejection letter did;
`
`however, state Amazon was willing to discuss making a production to Plaintiff.
`
`1.7
`
`Over the next three months, Plaintiff attempted to reach a resolution with the
`
`Company. Unfortunately, Amazon steadfastly insisted on including in a nondisclosure
`
`agreement draconian terms that would prevent Plaintiff from bringing certain of her claims
`
`derivatively and otherwise waiving her rights. Despite providing directly on point authority that
`
`Amazon's position was untenable and contrary to the law, it did not move. Further, Amazon
`
`never stated what documents it was willing to allow Plaintiff to inspect.
`
`1.8
`
`On February 1, 2021, Amazon stated that it would "follow up" with Plaintiff that
`
`week about her Inspection Demand. Instead, she has been met with silence. It has now been
`
`approximately six months since Plaintiff sent the Inspection Demand and she is apparently no
`
`closer to reviewing the demanded books and records. Amazon's actions have effectively denied
`
`Plaintiff her statutory rights.
`
`1.9
`
`In light of Amazon's effective refusal, Plaintiff now respectfully asks the Court
`
`to order Amazon to produce the demanded books and records that she is entitled to review as a
`
`stockholder of the Company.
`
`II.
`
`JURISDICTION AND VENUE
`
`2.1
`
`This Court retains general jurisdiction over each named defendant who is a
`
`resident of Washington. Additionally, this Court has specific jurisdiction over each named
`
`nonresident defendant because these defendants maintain sufficient minimum contacts with
`
`Washington to render jurisdiction by this Court permissible under traditional notions of fair play
`
`COMPLAINT - 3
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 4 of 46
`
`
`
`and substantial justice. Amazon is headquartered in Washington. Finally, exercising
`
`jurisdiction over any nonresident defendant is reasonable under these circumstances.
`
`2.2
`
`Venue is proper in this Court because defendant Amazon maintains executive
`
`offices in this County, a substantial portion of the transactions and wrongs complained of
`
`herein, including the defendant's primary participation in the wrongful acts detailed herein
`
`occurred in this County, and defendant has received substantial compensation in this County by
`
`doing business here and engaging in numerous activities that had an effect in this County.
`
`III. THE PARTIES
`
`3.1
`
`3.2
`
`Plaintiff Michele Rosati is an owner of Amazon's common stock.
`
`Defendant Amazon is a Delaware corporation with principal executive offices
`
`located at 410 Terry Avenue North, Seattle, Washington.
`
`IV.
`
`THE COMPANY'S UNAUTHORIZED COLLECTION OF INDIVIDUALS'
`INFORMATION VIOLATES THE LAW
`
`
`Biometrics and Facial Recognition Technology
`
`
`4.1
`
`Biometrics is the technical term for measurements used to identify people's
`
`unique physical characteristics. Examples of biometric identifiers include an individual's DNA,
`
`fingerprints, irises or retinas, voiceprints, and facial geometry. The uniqueness and potential
`
`permanence of biometric identifiers present an advantage for businesses to accurately identify
`
`and distinguish individuals. Businesses presently use biometrics in a wide variety of
`
`applications, including data collection.
`
`4.2
`
`One technological application of biometrics is facial recognition software.
`
`Facial recognition software uses biometrics to map facial features from a photograph or video.
`
`In particular, the software uses an algorithm that calculates a unique digital representation of the
`
`face based on the geometric relationship of a person's facial features (such as the distance
`
`COMPLAINT - 4
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 5 of 46
`
`
`
`between their eyes, ears, and nose), creating a face signature or map. The software then
`
`compares the information with a database of known faces to find a match.
`
`4.3
`
`Facial recognition technology has seen steady improvement over the past decade.
`
`Lower costs and increased accuracy have allowed companies such as Amazon to deploy
`
`increasingly sophisticated facial recognition software in their applications. However, this
`
`increased sophistication has raised serious privacy concerns. Biometrics present potential
`
`privacy threats to the individual if compromised, such as a heightened risk for identity theft.
`
`During a U.S. Senate hearing in 2012 on the use of facial recognition technology, Senator Al
`
`Franken noted that "[o]nce someone has your faceprint, they can get your name, they can find
`
`your social networking account, and they can find and track you in the street, in the stores that
`
`you visit, the Government buildings you enter, and the photos your friends post online." He
`
`added, "facial recognition technology can allow others to access all of that information from a
`
`distance, without your knowledge and in about as much time as it takes to snap a photo."
`
`Faceprints can even be used to identify protesters at political rallies and "target them for
`
`selective jailing and prosecution, stifling their First Amendment rights."
`
`4.4
`
`The U.S. Federal Trade Commission ("FTC") has also noted the public's
`
`concerns over privacy in social networks that "databases of photos or biometric data may be
`
`susceptible to breaches and hacking." The FTC urged companies using facial recognition
`
`technology to ask for consent before collecting biometric information from a photo. In its best
`
`practices guidelines, the FTC addressed social networks in particular, stating, "before using
`
`facial recognition to identify an individual it could not otherwise identify, the company should
`
`obtain the affirmative express consent of the individual in the image."
`
`/ / /
`
`
`COMPLAINT - 5
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 6 of 46
`
`
`
`The Illinois BIPA
`
`
`4.5
`
`In 2008, the Illinois General Assembly enacted the Illinois BIPA to enhance the
`
`state's "limited State law regulating the collection, use, safeguarding, and storage of
`
`biometrics[.]" 740 Ill. Comp. Stat. §14/5(e). BIPA defines a "biometric identifier" as including
`
`a "scan of hand or face geometry." 740 Ill. Comp. Stat. §14/10. The legislature noted that
`
`"[b]iometrics are unlike other unique identifiers that are used to access finances or other
`
`sensitive information," because while social security numbers can be changed if compromised,
`
`biometric data are "biologically unique to the individual," and "once compromised, the
`
`individual has no recourse, is at heightened risk for identity theft, and is likely to withdraw from
`
`biometric-facilitated transactions." 740 Ill. Comp. Stat. §14/5(c).
`
`4.6
`
`Under BIPA, companies must have a public, written policy establishing a
`
`retention schedule for biometric identifiers and information and guidelines for their permanent
`
`destruction. Moreover, a company may not collect or otherwise obtain a person or a customer's
`
`biometric identifier or biometric information without informing the subject in writing and
`
`securing a written release. Nor may a company profit from an individual's biometric identifiers
`
`and information. In particular, section 14/15(a)-(c) of BIPA provides:
`
`A private entity in possession of biometric identifiers or biometric
`a)
`information must develop a written policy, made available to the public,
`establishing a retention schedule and guidelines for permanently destroying
`biometric identifiers and biometric information when the initial purpose for
`collecting or obtaining such identifiers or information has been satisfied or within
`3 years of the individual's last interaction with the private entity, whichever
`occurs first. Absent a valid warrant or subpoena issued by a court of competent
`jurisdiction, a private entity in possession of biometric identifiers or biometric
`information must comply with its established retention schedule and destruction
`guidelines.
`
`No private entity may collect, capture, purchase, receive through trade, or
`b)
`otherwise obtain a person's or a customer's biometric identifier or biometric
`information, unless it first:
`
`COMPLAINT - 6
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 7 of 46
`
`1) informs the subject or the subject's legally authorized representative in
`writing that a biometric identifier or biometric information is being
`collected or stored;
`
`2) informs the subject or the subject's legally authorized representative in
`writing of the specific purpose and length of term for which a biometric
`identifier or biometric information is being collected, stored, and used; and
`
`
`
`
`
`
`
`
`
`3) receives a written release executed by the subject of the biometric
`identifier or biometric information or the subject's legally authorized
`representative.
`
`
`No private entity in possession of a biometric identifier or biometric
`c)
`information may sell, lease, trade, or otherwise profit from a person's or a
`customer's biometric identifier or biometric information.
`
`The Gender Shades Study
`
`4.7
`
`Facial recognition software algorithms that are trained with biased data can result
`
`in algorithmic discrimination.2 This, in turn, can lead to facial recognition products that are less
`
`effective at identifying certain types of faces. For example, an algorithm trained on dataset that
`
`underrepresents a group or subgroup (such as women or people of color) will have a higher rate
`
`of error in identifying members of those groups or subgroups.
`
`4.8
`
`In or around February 2018, researchers from the Massachusetts Institute of
`
`Technology and Microsoft Research released the Gender Shades study. In Gender Shades, the
`
`researchers analyzed three commercial facial recognition products and focused on each
`
`product's ability to accurately identify gender. They noted that prior studies had shown that
`
`"machine learning algorithms can discriminate based on classes like race and gender.
`
`
`2 See Buolamwini and Gebru, Gender Shades: Intersectional Accuracy Disparities in
`Commercial Gender Classification, Proceedings of Mach. Learning Research 81:1-15, at 1
`(2018), http://proceedings.mlr.press/v81/buolamwini18a/buolamwini18a .pdf ("Gender
`Shades").
`
`COMPLAINT - 7
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 8 of 46
`
`
`
`4.9
`
`The Gender Shades study determined that each product more accurately
`
`classified males than females, and light-skinned individuals than dark-skinned individuals. One
`
`product's error rate for classifying dark-skinned females was as high as 20.8%. The researchers
`
`concluded that "most improvement is needed on darker females specifically. More broadly, the
`
`error gaps between male and female classification along with lighter and darker classification
`
`should be closed."
`
`4.10
`
`In a follow up to Gender Shades, researchers examined the accuracy of
`
`Amazon's Rekognition biometric facial technology as compared to the technologies examined
`
`in the original study. The updated study found that Rekognition had an error rate of 31.37% for
`
`identifying dark-skinned females, as opposed to an error rate of 0% with respect to identifying
`
`light-skinned males.3
`
`4.11
`
`In recent years, an "arms race" has developed amongst for-profit companies
`
`seeking to become market leaders in the facial recognition arena. Critical to winning this battle
`
`has been the ability to claim a low identification error rate, in particular for these companies to
`
`herald the accuracy of their products, especially for identifying women and people of color.
`
`Flickr and IBM's Collection of Biometric Identifiers and Information
`
`4.12 Flickr, previously owned by Yahoo! Inc., is a photo-sharing website that had
`
`access to over 100 million photographs posted by its users. In or around 2014, Flickr compiled
`
`these photographs into a single dataset (the "Flickr Dataset"), and made it publicly available.
`
`However, Flickr did not inform or receive the consent of the individuals who uploaded the
`
`
`3 See Raji and Buolamwini, Actionable Auditing: Investigating the Impact of Publicly Naming
`Biased Performance Results of Commercial AI Products, Ass'n for the Advancement of
`Artificial Intelligence (2019), https://dam-prod.media.mit.edu/x/2019/01/24/AIES-
`19_paper_223.pdf.
`
`COMPLAINT - 8
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 9 of 46
`
`
`
`photographs or those who appeared in the photographs. Notably, the Flickr Dataset contained
`
`images of Illinois citizens and residents.
`
`4.13
`
`In the wake of the publication of the Gender Shades study, companies felt
`
`pressured to improve their accuracy of their facial recognition products and reduce the bias
`
`therein. In or around January 2019, IBM announced the release of its Diversity in Facesa
`
`new dataset consisting of one million images culled from the Flickr Datasetfor the purpose of
`
`improving the ability of facial recognition systems to fairly and accurately identify all
`
`individuals (the "Diversity in Faces Dataset.").4
`
`4.14
`
`In creating the Diversity in Faces Dataset, IBM scanned the facial geometry of
`
`each image and created a "comprehensive set of annotations of intrinsic facial features that
`
`includes craniofacial distances, areas and ratios, facial symmetry and contrast, skin color, age
`
`and gender predictions, subjective annotations, and pose and resolution." Notably, IBM did not
`
`seek or receive permission from those who uploaded their photographs to Flickr to include their
`
`images in the Diversity in Faces Dataset, let alone to perform scans of their facial geometries or
`
`obtain or profit from their biometric identifiers and information. In or around April 2019, IBM
`
`published a research report describing its Diversity in Faces Dataset and making clear that the
`
`dataset contained the biometric identifiers and information of each individual who appeared
`
`therein.5
`
`
`4 See John R. Smith, IBM Research Releases "Diversity in Faces" Dataset to Advance Study of
`Fairness in Facial Recognition Systems, IBM (Jan. 29, 2019), https://www.ibm.com/blogs/
`research/2019/01/diversity-in-faces/.
`
`5 Michele Merler, et al., Diversity in Faces, IBM Research AI (Apr. 10, 2019),
`https://arxiv.org/pdf/1901.10436.pdf.
`
`
`COMPLAINT - 9
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 10 of 46
`
`
`
`4.15
`
`In addition, IBM made the Diversity in Faces Dataset available to other for-profit
`
`companies that developed, produced, marketed, sold, or otherwise used facial recognition
`
`products and technologies in connection with their businesses. Once granted access by IBM,
`
`the company seeking access had to download the Diversity in Faces Dataset through a link
`
`provided by IBM. This downloaded information included the biometric identifiers and
`
`information extracted from each photograph in the dataset, and links to each photograph on
`
`Flickr. From the Flickr links, the companies were able to identify the Flickr user who uploaded
`
`the photograph to Flickr, that user's homepage and other posted material, and each photograph's
`
`metadata, including any available geo-tags relating to where the photograph was taken or
`
`uploaded.
`
`Amazon Collects and Stores Biometric Data from Its Users
`
`4.16 Amazon's core facial recognition product is Rekognition, which launched in
`
`November 2016. Rekognition allows users to match new images of faces with existing, known
`
`facial images "based on their visual geometry, including the relationship between the eyes, nose,
`
`brow, mouth, and other facial features." Rekognition is a cornerstone of many of Amazon's
`
`largest consumer products and services, including its photo platform, Amazon Photosits smart
`
`home systems and cameras, and its virtual assistant technology, Alexa.
`
`4.17 Amazon is also the largest provider of facial recognition technology to law
`
`enforcement agencies. The Company has marketed its Rekognition software to agencies such
`
`as the U.S. Immigration and Customs Enforcement and the Federal Bureau of Investigation, to
`
`monitor individuals they consider "people of interest." Amazon has also partnered with more
`
`than 1,300 law enforcement agencies, allowing them to use footage from their Ring home
`
`security cameras in criminal investigations. Amazon has expanded these efforts marketing their
`
`COMPLAINT - 10
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 11 of 46
`
`
`
`facial recognition software to government agencies despite warnings from consumers,
`
`employees, members of Congress, and stockholders.
`
`4.18
`
`In July 2018, the American Civil Liberties Union of Northern California
`
`("ACLU") published the results of a study it conducted regarding Rekognition's accuracy.6
`
`According to the study, Rekognition incorrectly matched twenty-eight members of the U.S.
`
`Congress to people who had been arrested for a crime. The false matches disproportionately
`
`involved people of color. That summer, nearly seventy civil rights and research organizations
`
`wrote a letter to Amazon's Chief Executive Officer, Jeffrey P. Bezos ("Bezos"), demanding that
`
`Amazon stop providing facial recognition technology to governments. In their letter, they
`
`called the Company to "stand up for civil rights and civil liberties," stating "Rekognition is a
`
`powerful surveillance system readily available to violate rights and target communities of
`
`color." Amazon's own employees demanded the Company to stop selling its Rekognition facial
`
`recognition software to law enforcement, citing concerns over the "unique threat to civil rights
`
`and especially to the immigrants and people of color under attack by [President Donald J.
`
`Trump's] administration."7
`
`4.19 Seeking to improve the accuracy of its facial recognition products and
`
`technologies, Amazon allegedly obtained IBM's Diversity in Faces Dataset after IBM made it
`
`available to for-profit companies in early 2019. To do so, Amazon used the links provided by
`
`IBM to download or otherwise obtain from the Flickr Dataset each photograph in order to
`
`
`6 Jacob Snow, Amazon's Face Recognition Falsely Matched 28 Members of Congress with
`Mugshots, ACLU.org (July 26, 2018), https://www.aclu.org/blog/privacy-
`technology/surveillance-technologies/amazons-face-recognition-falsely-matched-28.
`
`7 Kate Conger, Amazon Workers Demand Jeff Bezos Cancel Face Recognition Contracts with
`Law Enforcement, GIZMODO (June 21, 2018), https://gizmodo.com/amazon-workers-demand-
`jeff-bezos-cancel-face-recognitio-1827037509.
`
`COMPLAINT - 11
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 12 of 46
`
`
`
`associate the biometric identifiers and information provided by IBM with the actual
`
`photographs to which the biometric data related. Amazon's collection and use of the Diversity
`
`in Faces Dataset allowed it to profit from such data by allowing Amazon to improve the
`
`effectiveness of its own facial recognition technology and products.
`
`Amazon's Collection and Storing of Biometric Data Violates BIPA
`
`4.20
`
`In direct violation of BIPA, the Company stored its users' biometric information
`
`without informing them or securing their written consent. The Company has also failed to
`
`develop a written policy, made available to the public, establishing a retention schedule and
`
`guidelines for users to permanently destroy biometric identifiers when the initial purpose for
`
`collection has been satisfied. As such, the Company has violated the express language of BIPA.
`
`4.21 These violations have exposed the Company to substantial harm. On July 14,
`
`2020, a federal consumer class action lawsuit was filed on behalf of Amazon users in the U.S.
`
`District Court for the Western District of Washington alleging that Amazon unlawfully obtained
`
`and stored their biometrics information and identifiers (the "Consumer Class Action"). The
`
`Consumer Class Action asserted causes of action under section 14/15(b)-(c) of Chapter 740 of
`
`the Illinois Compiled Statutes. Specifically, the plaintiffs asserted that Amazon never informed
`
`them, by written notice or otherwise, that Amazon collected, stored, and used their biometric
`
`identifiers and information, or of the specific purpose and length of term for which their
`
`biometric identifiers were being collected, stored, and used. Nevertheless, when the plaintiffs
`
`uploaded photos to their accounts, Amazon extracted from those photos their biometric
`
`identifiers and stored them in its databases.
`
`/ / /
`
`/ / /
`
`COMPLAINT - 12
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 13 of 46
`
`
`
`4.22
`
`In a similar action concerning BIPA violations, Facebook, Inc. had to pay $650
`
`million to settle the matter.8
`
`V.
`
`THE COMPANY'S ANTICOMPETITIVE PRACTICES
`
`5.1
`
`Plaintiff is also seeking to investigate whether Amazon's fiduciaries authorized
`
`or allowed the Company to engage in anticompetitive practices, leading to U.S. and
`
`international regulatory scrutiny. Amazon is one of the world's largest online marketplaces.
`
`Amazon has a dual role as an online platform: it sells products on its website as a retailer and
`
`also provides a marketplace where independent sellers can sell products directly to consumers.
`
`No other U.S. retailer operates a marketplace even close to the size of Amazon's. Over the past
`
`few years, Amazon's market share in U.S. online commerce has increased to about 40%, which
`
`is about seven times more than the next competitor.
`
`5.2
`
`The Company makes and sells its own products to compete with brand names on
`
`its own platform. Amazon's private-label business encompasses more than forty-five brands.
`
`5.3
`
`The Company currently faces significant regulatory inquiries into its practices,
`
`specifically over whether it unfairly uses its size and platform against competitors and other
`
`sellers on its site. Amazon disputes that it abuses its power and size.
`
`The European Union Investigates the Company's Anticompetitive Practices
`
`5.4
`
`In or around September 2018, the European Union's ("EU") top antitrust
`
`enforcer, the European Commission, began examining whether Amazon abused its dual role as
`
`a seller of its own products and a marketplace operator.9 The European Commission also
`
`
`8 Patel v. Facebook, Inc., No. 3:15-cv-03747 (N.D. Cal.)
`
`9 This fact-finding mission stems from the European Commission's e-commerce sector inquiry
`into business practices that may restrict competition, beginning in May 2015. The European
`Commission published its findings from the initial inquiry in May 2017.
`
`COMPLAINT - 13
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 14 of 46
`
`
`
`examined whether the Company was gaining a competitive advantage from the data it gathers
`
`on third-party sellers, as Amazon continuously collects data about sellers' activity on its
`
`platform.
`
`5.5
`
`In September 2019, the European Commission opened a formal antitrust
`
`investigation to assess whether Amazon's use of sensitive data from independent sellers on its
`
`marketplace breached EU competition rules. The EU has strict competition rules on
`
`anticompetitive agreements between companies and on the abuse of a dominant position.
`
`Commissioner Margrethe Vestager, in charge of competition policy, stated:
`
`European consumers are increasingly shopping online. E-commerce has boosted
`retail competition and brought more choice and better prices. We need to ensure
`that large online platforms don't eliminate these benefits through anti-competitive
`behaviour. I have therefore decided to take a very close look at Amazon's
`business practices and its dual role as marketplace and retailer, to assess its
`compliance with EU competition rules.
`
`5.6
`
`As part of its investigation, the European Commission looked into the standard
`
`agreements between Amazon and marketplace sellers, which allow Amazon's retail business to
`
`analyze and use third-party seller data. In particular, the European Commission focused on
`
`whether and how the use of accumulated marketplace seller data by Amazon as a retailer affects
`
`competition. Based on the European Commission's preliminary fact-finding, Amazon appeared
`
`to use competitively sensitive information about marketplace sellers, their products, and
`
`transactions on the marketplace.
`
`5.7
`
`Recently, reports have surfaced that Amazon faces EU antitrust charges over its
`
`use of third-party seller data.10 The charges will reportedly accuse Amazon of using data
`
`
`10 See Jon Porter, Amazon Reportedly Faces EU Antitrust Charges over Use of Third-Party
`Seller Data, The Verge (June 11, 2020),
`https://www.theverge.com/2020/6/11/21287672/amazon-european-union-antitrust-charges-
`third-party-seller-data.
`
`COMPLAINT - 14
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 15 of 46
`
`
`
`gathered from sellers on its marketplace to compete against them. Notably, Commissioner
`
`Vestager has previously fined major U.S. tech giants, including Google, Qualcomm, and
`
`Facebook, for their anticompetitive practices in Europe.
`
`The Wall Street Journal Exposes Amazon's Improper Use of Third-Party Seller Data
`
`5.8
`
`On April 23, 2020, the Wall Street Journal (the "WSJ") published an article titled
`
`"Amazon Scooped Up Data from Its Own Sellers to Launch Competing Products." The article
`
`described how the Company used data about independent sellers on its platform to develop
`
`competing products, in violation of its own policies. Such proprietary information can help
`
`Amazon decide how to price an item, which features to copy, or whether to enter a product
`
`segment based on its earning potential.
`
`5.9
`
`The Company has claimed publicly that "we strictly prohibit our employees from
`
`using nonpublic, seller-specific data to determine which private label products to launch."
`
`Further, Amazon's associate general counsel has told Congress, "[w]e don't use individual seller
`
`data directly to compete" with businesses on the Company's platform.
`
`5.10
`
`In violation of these internal policies, however, the Company's employees used
`
`the collected proprietary information for Amazon's own benefit. Although Amazon has stated it
`
`has restrictions in place to keep its private-label executives from accessing data on specific
`
`sellers in its marketplace, former employees admitted those rules were not uniformly enforced.
`
`In fact, according to some former employees, using such data was a common practice that was
`
`discussed openly in meetings they attended. Former executives said they were told frequently
`
`by management that Amazon brands should make up more than 10% of retail sales by 2022.
`
`Managers of private-label product categories were told to create $1 billion businesses for their
`
`segments.
`
`COMPLAINT - 15
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`
`Case 2:21-cv-00409 Document 1-2 Filed 03/26/21 Page 16 of 46
`
`
`
`5.11 According to the WSJ, Company employees used the third-party sellers' data to
`
`launch and benefit Amazon products. Some executives used proprietary information to research
`
`best-selling items they might want to compete against. If access was restricted, managers would
`
`ask an Amazon business analyst to create reports featuring the information.
`
`5.12 For instance, Amazon's employees accessed documents and data about a best-
`
`selling car-trunk organizer sold by a third-party vendor called Fortem. Fortem i