`
`
`
`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
`
`
`
`
`IN THE UNITED STATES DISTRICT COURT
`FOR THE WESTERN DISTRICT OF WASHINGTON AT SEATTLE
`
`STEVEN VANCE and TIM JANECYK, for
`themselves and others similarly situated,
`
`
`Plaintiffs,
`
`
`
`
`
`
`
`
`
`
`
`
`
`No. ________________
`
`CLASS ACTION COMPLAINT
`
`JURY DEMAND
`
`v.
`
`
`
`
`
`
`
`
`MICROSOFT CORPORATION,
`
`
`Defendant.
`
`
`
`
`CLASS ACTION COMPLAINT
`
`Plaintiffs STEVEN VANCE and TIM JANECYK, on behalf of themselves and all other
`
`similarly situated individuals (“Plaintiffs”), by and through their respective attorneys, bring this
`
`Class Action Complaint against Defendant Microsoft Corporation (“Microsoft”) and allege the
`
`following:
`
`INTRODUCTION
`
`1.
`
`Facial recognition technology – once a thing only seen in movies – now
`
`threatens to end individual privacy. Public and private entities increasingly deploy facial
`
`recognition products to determine a private citizens’ identities, as well as other personal
`
`information, such as their addresses, phone numbers, whereabouts and acquaintances.
`
`2.
`
`Unlike the way facial recognition technology is depicted in the movies, the
`
`actual technology is plagued by a major problem – it is inaccurate, especially when it comes to
`
`correctly identifying women and people of color.
`
`3.
`
`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
`
`CLASS ACTION COMPLAINT - 1
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 2 of 26
`
`
`
`has been to the ability to claim a low identification error rate – i.e., the for-profit companies
`
`want to herald the accuracy of their products, including accuracy in identifying woman and
`
`people of color.
`
`4.
`
`In its effort to improve its facial recognition technology, Defendant Microsoft
`
`violated Illinois’ Biometric Information Privacy Act, 740 ILCS 14/1, et seq. (“BIPA”), by,
`
`among other things, unlawfully collecting, obtaining, storing, using, possessing and profiting
`
`from the biometric identifiers and information of Plaintiffs Vance and Janecyk and all other
`
`similarly situated Illinois residents and citizens (hereinafter, the “Class Members”).
`
`5.
`
`Plaintiffs bring this Class Action Complaint seeking: (a) statutory damages of
`
`$5,000 per BIPA violation, or, alternatively, if Defendant Microsoft acted negligently, $1,000
`
`per BIPA violation, along with attorneys’ fees and costs; (b) disgorgement of Defendant’s ill-
`
`gotten gains derived from the use of the unlawfully-acquired data; and (c) an injunction (i)
`
`barring Defendant from any further use of Illinois citizens’ and residents’ biometric identifiers
`
`and information; (ii) barring Defendant from continuing to collect, obtain, store, use, possess
`
`and profit from Plaintiffs’ and Class Members’ biometric identifiers and information; and (iii)
`
`requiring Defendant to delete and destroy Plaintiffs’ and Class Members’ biometric identifiers
`
`and information.
`
`PARTIES
`
`6.
`
`At relevant times, Plaintiff STEVEN VANCE was – and remains – an Illinois
`
`resident who lived in the Northern District of Illinois. Defendant Microsoft collected, obtained,
`
`stored, used, possessed and profited from Plaintiff Vance’s biometric identifiers and
`
`information – namely, facial geometric scans of Plaintiff Vance.
`
`CLASS ACTION COMPLAINT - 2
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 3 of 26
`
`
`
`7.
`
`At relevant times, Plaintiff TIM JANECYK was – and remains – an Illinois
`
`resident who lived in the Northern District of Illinois. Defendant Microsoft collected, obtained,
`
`stored, used, possessed and profited from Plaintiff Janecyk’s biometric identifiers and
`
`information – namely, facial geometric scans of Plaintiff Janecyk.
`
`8.
`
`Defendant Microsoft
`
`is a Washington corporation based
`
`in Redmond,
`
`Washington.
`
`JURISDICTION AND VENUE
`
`9.
`
`This Court has jurisdiction pursuant to 28 U.S.C. § 1332(d)(2) (the “Class Action
`
`Fairness Act”) because sufficient diversity of citizenship exists between the parties in this action,
`
`the aggregate amount in controversy exceeds $5,000,000, exclusive of interests and costs, and
`
`there are 100 or more members of the Class. Because it is estimated that the Class will have
`
`thousands of members and Defendant Microsoft’s intentional and reckless violations of BIPA
`
`are punishable by statutory damages of $5,000 per violation, the amount in controversy is well
`
`in excess of $5,000,000. This Court has supplemental jurisdiction over the state law claim
`
`pursuant to 28 U.S.C. § 1367.
`
`10.
`
`This Court has personal jurisdiction over Defendant Microsoft because it is at
`
`home in the Western District of Washington. As alleged above, Microsoft is a Washington
`
`corporation headquartered in Redmond, Washington.
`
`11.
`
`Venue is proper under 28 U.S.C. § 1391(b)(1) because Defendant Microsoft
`
`resides in the Western District of Washington.
`
`
`
`
`
`CLASS ACTION COMPLAINT - 3
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 4 of 26
`
`
`
`Biometric Identifiers
`
`FACTUAL ALLEGATIONS
`
`12.
`
`Every individual has unique features by which he or she can be identified using a
`
`set of standard quantitative measurements, commonly referred to as “biometric identifiers.”
`
`13.
`
`For example, the shape of and distance between tiny ridges on each person’s
`
`finger are unique, so measures of those features can be used to identify a specific individual as
`
`the person who made a fingerprint.
`
`14.
`
`Each person also has a unique facial geometry composed of, among other
`
`measures, distances between key facial landmarks and ratios between those distances.
`
`15.
`
`Once a picture of a person’s face is scanned and its biometric measurements are
`
`captured, computers can store that information and use it to identify that individual any other
`
`time that person’s face appears on the internet, in a scanned picture or footage from any of the
`
`billions of cameras that are constantly monitoring the public’s daily lives.
`
`16.
`
`Unlike fingerprints, however, facial biometrics are readily observable and, thus,
`
`present a grave and immediate danger to privacy, individual autonomy and liberty.
`
`The Illinois Biometric Information Privacy Act
`
`17.
`
`Through BIPA, Illinois strictly regulates the collection, obtainment, storage, and
`
`use of biometric identifiers and information.
`
`18.
`
`Under BIPA, biometric identifiers include a scan of an individual’s face
`
`geometry. 740 ILCS § 14/10.
`
`19.
`
`Under BIPA, biometric information is “any information . . . based on an
`
`individual’s biometric identifier used to identify an individual.” 740 ILCS § 14/10.
`
`CLASS ACTION COMPLAINT - 4
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 5 of 26
`
`
`
`20.
`
`According to the Illinois General Assembly: “[b]iometrics are unlike other
`
`unique identifiers that are used to access finances or other sensitive information. For example,
`
`social security numbers, when compromised, can be changed. Biometrics, however, are
`
`biologically unique to the individual; therefore, once compromised, the individual has no
`
`recourse, is at heightened risk for identity theft, and is likely to withdraw from biometric-
`
`facilitated transactions.” 740 ILCS § 14/5(c).
`
`21.
`
`Pursuant to BIPA, a private entity is, among other things: (a) prohibited from
`
`collecting or otherwise obtaining an individual’s biometric identifiers and information without
`
`providing written notice and obtaining a written release; (b) prohibited from profiting from an
`
`individual’s biometric identifiers and information; and (c) required, to the extent it is in
`
`possession of biometric identifiers or information, to develop a written policy, made available to
`
`the public, that establishes a retention schedule and guidelines for permanently destroying such
`
`identifiers and information. 740 ILCS § 14/15.
`
`22.
`
`BIPA provides for a private right of action and allows a prevailing party to
`
`recover liquidated damages in the amount of: (a) $1,000 or actual damages, whichever is greater,
`
`for negligent violations of its provisions; and (b) $5,000 or actual damages, whichever is greater,
`
`for intentional or reckless violations of its provisions. 740 ILCS § 14/20. BIPA also allows for
`
`the recovery of attorneys’ fees and costs and injunctive relief. 740 ILCS § 14/20.
`
`Facial Recognition Technology
`
`23.
`
`Facial recognition is a form of computer artificial intelligence, the goal of which
`
`is to “create systems that detect, recognize, verify and understand characteristics of human
`
`faces.”1
`
`
`
`1 Michele Merler, et al., Diversity in Faces, IBM Research AI (Apr. 10, 2019) (“Diversity in Faces”).
`CLASS ACTION COMPLAINT - 5
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 6 of 26
`
`
`
`24.
`
`To do this well, the algorithms driving facial recognition technology must be
`
`trained with and fed vast quantities of images of a diverse array of faces. To satisfy the ever-
`
`growing demand for myriad high-resolution images of faces, unchecked companies have begun
`
`turning to the internet, where photographs are sometimes taken without the photographer’s or
`
`subject’s knowledge or consent. This has been called the dirty little secret of AI training sets.
`
`Researchers often just grab whatever images they can find “in the wild.”
`
`25.
`
`Facial recognition products rely on machine learning algorithms that are trained
`
`with labeled data.2 As a result, algorithms trained with biased data can result in algorithmic
`
`discrimination3 which, in turn, can lead to facial recognition products that are less effective at
`
`identifying certain types of faces.
`
`26.
`
`For example, an algorithm trained on a dataset that underrepresents a group or
`
`subgroup – e.g., woman or people of color – will have a higher rate of error with respect to
`
`identifying members of those groups or subgroups.
`
`27.
`
`Historically, available datasets on which facial recognition algorithms were
`
`trained contained a disproportionate number of light-skinned males.
`
`Flickr
`
`28.
`
`At relevant times, Flickr was a photo-sharing website that had access to over 100
`
`million photographs posted by Flickr users.
`
`29.
`
`In or about 2014, Flickr – through its parent company Yahoo! – compiled
`
`approximately 100 million Flickr photographs into a single dataset (the “Flickr Dataset”) and
`
`made the dataset publicly available.
`
`
`
`2 Joy Buolamwini, et al., Gender Shades: Intersectional Accuracy Disparities in Commercial Gender
`Classification, Proceedings of Machine Learning Research 81:1-15 (2018) at 1.
`3 Id.
`CLASS ACTION COMPLAINT - 6
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 7 of 26
`
`
`
`30.
`
`Flickr did so without informing or receiving the consent of the individuals who
`
`uploaded these photographs to Flickr or who appeared in these photographs.
`
`31.
`
`Flickr contended that its purpose in releasing the Flickr Dataset was to help
`
`improve the accuracy and reliability of facial recognition technology.
`
`32.
`
`The Flickr Dataset contained images of Illinois citizens and residents, including
`
`images of Plaintiffs and Class Members.
`
`The Gender Shades Study
`
`33.
`
`In or about February 2018, researchers released Gender Shades: Intersectional
`
`Accuracy Disparities in Commercial Gender Classification (“Gender Shades”) in which they
`
`noted that prior studies had shown that “machine learning algorithms can discriminate based on
`
`classes like race and gender.”4
`
`34.
`
`Building on that prior research, the researchers analyzed three commercial facial
`
`recognition products – including a Microsoft product – focusing on each product’s ability to
`
`accurately identify gender.5
`
`35.
`
`The study determined that each product more accurately classified: (a) males
`
`than females; and (b) lighter individuals than darker individuals.6
`
`36.
`
`Significantly, the error rate with respect to accurately classifying darker females
`
`was 20.8% for Defendant Microsoft, specifically, and as high as approximately 34.7%.7
`
`
`
`
`
`
`
`4 Id.
`5 See id. at 8.
`6 Id.
`7 Id. at 9.
`
`CLASS ACTION COMPLAINT - 7
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 8 of 26
`
`
`
`37.
`
`The researchers concluded that the “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.”8
`
`Response to Gender Shades
`
`38.
`
`In the aftermath of Gender Shades, companies felt pressured to improve the
`
`accuracy of, and reduce the bias in, their facial recognition products.
`
`39.
`
`In or about April 2019, International Business Machines Corporation (“IBM”)
`
`noted that a “critical aspect limiting face recognition performance in practice is facial diversity,”
`
`begging the question “does the training data for [face recognition] systems fairly represent the
`
`distribution of faces we see in the world?”9
`
`40.
`
`To respond to the issue, IBM created Diversity in Faces – a new dataset
`
`consisting of one million images culled from the Flickr Dataset – for the purpose of improving
`
`the ability of facial recognition systems to fairly and accurately identify all individuals (the
`
`“Diversity in Faces Dataset”).10
`
`41.
`
`In creating the Diversity in Faces Dataset, IBM scanned the facial geometry of
`
`each image contained in the dataset 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.”11
`
`
`
`8 Id. at 11.
`9 Diversity in Faces, supra, at 1.
`10 See id.
`11 Id. at 2.
`
`CLASS ACTION COMPLAINT - 8
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 9 of 26
`
`
`
`42.
`
`To build the Diversity in Faces Database, IBM extracted 19 facial landmark
`
`points from each image in the dataset to determine 68 key points for each face.12
`
`43.
`
`IBM used the 19 facial landmark points to extract craniofacial features for each
`
`image, as shown in the figure below13:
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`44.
`
`The Diversity in Faces Dataset contained the biometric identifiers and
`
`information of Plaintiffs and Class Members.
`
`45.
`
`IBM did not seek nor receive permission from Plaintiffs or Class Members to
`
`include their images in the Diversity in Faces Dataset, let alone to perform scans of their facial
`
`geometries or to otherwise collect, obtain, store, use, possess or profit from their biometric
`
`identifiers and information.
`
`
`
`12 Id. at 9.
`13 Id. at 9-10.
`CLASS ACTION COMPLAINT - 9
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`Case 2:20-cv-01082 Document 1 Filed 07/14/20 Page 10 of 26
`
`
`
`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
`
`
`
`46.
`
`In or about April 2019, IBM published a journal article describing the Diversity
`
`in Faces Dataset in great detail and making clear that the dataset contained the biometric
`
`identifiers and information of each individual who appeared in the dataset.
`
`47.
`
`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 for-profit businesses.14
`
`48.
`
`To obtain the Diversity in Faces Dataset from IBM, a company had to apply for
`
`permission from IBM via an online questionnaire.
`
`49.
`
`If IBM granted access to the Diversity in Faces Dataset, the company seeking
`
`access had to download the dataset from a link provided by IBM.
`
`50.
`
`The information provided to companies that downloaded the Diversity in Faces
`
`Dataset included the biometric identifiers and information extracted from each photograph in
`
`the dataset and links to each photograph on Flickr from which IBM extracted the biometric data.
`
`51.
`
`From the Flickr links IBM provided to companies that downloaded the Diversity
`
`in Faces Dataset, the companies were able to identify the Flickr user who uploaded the
`
`photograph to Flickr, view the Flickr user’s homepage and other posted material, and view each
`
`photograph’s metadata, including any available geo-tags relating to where the photograph was
`
`taken or uploaded.
`
`Defendant Microsoft Obtained the Diversity in Faces Dataset
`
`52.
`
`At relevant times, Defendant Microsoft developed, produced, marketed and
`
`otherwise used facial recognition products and technologies in connection with its business.
`
`14 See Diversity in Faces, supra.
`
`
`
`
`
`CLASS ACTION COMPLAINT - 10
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 11 of 26
`
`
`
`53.
`
`Among Defendant Microsoft’s facial recognition products were its Cognitive
`
`Services Face Application Program Interface and its Face Artificial Intelligence service that
`
`allowed customers to embed facial recognition into their apps without having to have any
`
`machine learning expertise.
`
`54.
`
`At relevant times, Defendant Microsoft was aware of the importance of
`
`companies following the rule of law when it comes to facial recognition technology and urged
`
`governments to regulate the technology.15
`
`55.
`
`After IBM made the Diversity in Faces Dataset available, Defendant Microsoft
`
`applied for and obtained the Diversity in Faces Dataset from IBM.
`
`56.
`
`On information and belief, upon obtaining the Diversity in Faces Dataset from
`
`IBM, Defendant Microsoft used the links provided by IBM to download, copy or otherwise
`
`obtain from Flickr each photograph in the dataset, including Plaintiffs’ photographs, in order to
`
`associate the biometric identifiers and information provided by IBM with the actual
`
`photographs to which the biometric data related.
`
`57.
`
`Defendant Microsoft obtained the Diversity in Faces Dataset in order to improve
`
`the fairness and accuracy of its facial recognition products and technologies.
`
`58.
`
`Defendant Microsoft profited from the biometric identifiers and information
`
`contained in the Diversity in Faces Dataset because those biometric identifiers and information
`
`allowed Microsoft to improve its facial recognition products and technologies, including, upon
`
`information and belief, by allowing Microsoft to improve the effectiveness of its facial
`
`
`
`15 Brad Smith, Facial Recognition: It’s Time for Action, Microsoft on the Issues (Dec. 6, 2018),
`https://blogs.microsoft.com/on-the-issues/2018/12/06/facial-recognition-its-time-for-action/ (last
`accessed on Mar. 23, 2020).
`
`CLASS ACTION COMPLAINT - 11
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 12 of 26
`
`
`
`recognition technology on a diverse array of faces, thereby making those products and
`
`technologies more valuable in the commercial marketplace.
`
`59.
`
`Defendant Microsoft conducted extensive business within Illinois related to the
`
`facial recognition products it unlawfully developed using Plaintiffs’ and Class Members’
`
`biometric identifiers and information, including: (i) selling its facial recognition products to
`
`third-party clients through an Illinois-based vendor; (ii) working closely with an Illinois-based
`
`business to build new applications for its facial recognition technology; and (iii) working with
`
`the University of Illinois, among others, to build and promote a “digital transformation institute”
`
`aimed at “accelerating the application of artificial intelligence” throughout business and society.
`
`As such, Illinois had and has a direct interest in regulating the unlawful conduct alleged herein
`
`in order to protect the rights and interests of its residents and citizens.
`
`Allegations Related to Plaintiff Vance
`
`60.
`
`In or about 2008, Plaintiff Vance uploaded to Flickr from his computer in Illinois
`
`a photograph of himself and two family members (the “2008 Photo”).
`
`61.
`
`In addition to the 2008 Photo, Plaintiff Vance uploaded numerous other
`
`photographs to Flickr.
`
`62.
`
`At relevant times, Plaintiff Vance’s publicly-accessible Flickr profile page
`
`clearly identified his Chicago, Illinois residence and provided a method for those accessing his
`
`page to contact him directly via Flickr’s internal “FlickrMail” direct message system, which
`
`Defendant Microsoft chose not to do.
`
`63.
`
`The 2008 Photo, as well as numerous other photographs uploaded to Flickr by
`
`Plaintiff Vance, were part of the Diversity in Faces Dataset obtained by Defendant Microsoft.
`
`CLASS ACTION COMPLAINT - 12
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 13 of 26
`
`
`
`64.
`
`Based on the links Defendant Microsoft received from IBM, at relevant times, it
`
`knew that each of Plaintiff Vance’s photographs in the Diversity in Faces Dataset – including
`
`the 2008 Photo – originated from, and was affiliated with, his Flickr account.
`
`65.
`
`Defendant Microsoft never advised or informed Plaintiff Vance or his legal
`
`authorized representative in writing: (a) that it collected, stored and used Plaintiff Vance’s
`
`biometric identifiers and information; or (b) of the specific purpose and length of term for
`
`which Plaintiff Vance’s biometric identifiers and information were being collected, stored and
`
`used.
`
`66.
`
`Defendant Microsoft never received a written release executed by Plaintiff
`
`Vance or his legally authorized representative to collect, capture, receive, obtain, store or use
`
`his biometric identifiers and information.
`
`67.
`
`As alleged in more detail below, Defendant Microsoft’s conduct has injured
`
`Plaintiff Vance and subjected him to additional imminent and certainly impending injuries.
`
`Allegations Related to Plaintiff Janecyk
`
`68.
`
`Plaintiff Janecyk is an accomplished photographer, having focused his work in
`
`portraiture and street life photography.
`
`69.
`
`In 2008, Plaintiff Janecyk signed up for a Flickr account in the Village of Tinley
`
`Park, Illinois, and has since then uploaded in excess of a thousand of his photographs to Flickr.
`
`Among those photographs is a 2011 photograph depicting Plaintiff Janecyk’s own face (the
`
`“2011 Photo”), which Plaintiff Janecyk uploaded to Flickr from his device in Illinois.
`
`70.
`
`At relevant times, Plaintiff Janecyk’s publicly-accessible Flickr profile page
`
`clearly identified his Illinois residence and provided a method for those accessing his page to
`
`CLASS ACTION COMPLAINT - 13
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`Case 2:20-cv-01082 Document 1 Filed 07/14/20 Page 14 of 26
`
`
`
`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
`
`
`
`contact him directly via Flickr’s internal “FlickrMail” direct message system, which Defendant
`
`Microsoft chose not to do.
`
`71.
`
`The 2011 Photo, as well as numerous other photographs uploaded to Flickr by
`
`Plaintiff Janecyk, were part of the Diversity in Faces Dataset obtained by Defendant Microsoft.
`
`72.
`
`Based on the links Defendant Microsoft received from IBM, at relevant times, it
`
`knew that each of Plaintiff Janecyk’s photographs in the Diversity in Faces Dataset – including
`
`the 2011 Photo – originated from, and was affiliated with, his Flickr account.
`
`73.
`
`Defendant Microsoft never advised or informed Plaintiff Janecyk or his legal
`
`authorized representative in writing: (a) that it collected, stored and used Plaintiff Janecyk’s
`
`biometric identifiers and information; or (b) of the specific purpose and length of term for
`
`which Plaintiff Janecyk’s biometric identifiers and information were being collected, stored and
`
`used.
`
`74.
`
`Defendant Microsoft never received a written release executed by Plaintiff
`
`Janecyk or his legally authorized representative to obtain, collect, store or use his biometric
`
`identifiers and information.
`
`75.
`
`As alleged in more detail below, Defendant Microsoft’s conduct has injured
`
`Plaintiff Janecyk and subjected him to additional imminent and certainly impending injuries.
`
`Plaintiffs’ and Class Members’ Injuries and Damages
`
`76.
`
`As a result of Defendant Microsoft’s unlawful conduct, Plaintiffs and Class
`
`Members have already sustained injuries and face many more imminent and certainly
`
`impending injuries, which injuries they will continue to suffer.
`
`77.
`
`Defendant Microsoft chose to use and profit from biometric identifiers and
`
`information scanned from photographs that were uploaded from Illinois; managed via Illinois-
`
`CLASS ACTION COMPLAINT - 14
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`Case 2:20-cv-01082 Document 1 Filed 07/14/20 Page 15 of 26
`
`
`
`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
`
`
`
`based user accounts, computers and mobile devices; and/or created in Illinois. In so doing,
`
`Microsoft exposed Illinois residents and citizens to ongoing privacy risks within Illinois,
`
`knowing that its conduct would injure those residents and citizens within Illinois. Moreover,
`
`Microsoft knew or had reason to know that obtaining Illinois residents’ and citizens’ biometric
`
`identifiers and information in violation of BIPA would deprive those residents and citizens of
`
`their statutorily-protected privacy rights, neutralize Illinois residents’ and citizens’ abilities to
`
`control access to their biometric identifiers and information via their Illinois-managed devices,
`
`expose Illinois residents and citizens to potential surveillance and other privacy harms as they
`
`went about their lives within the state and deter Plaintiffs and Class Members from publicly
`
`posting photographs. As such, Illinois had and has a direct interest in regulating the unlawful
`
`conduct alleged herein in order to protect the rights and interests of its residents and citizens.
`
`78.
`
`As the Illinois General Assembly has found and the Illinois Supreme Court has
`
`confirmed, the harm to Plaintiffs and Class Members as a result of Defendant Microsoft’s
`
`unlawful conduct has already occurred.
`
`79.
`
`Further, as businesses worldwide compete to develop ever more advanced facial
`
`recognition technology, the race for data imperils the privacy of individuals everywhere,
`
`including the privacy of Plaintiffs and Class Members. Public policy in Illinois provides that
`
`given the risks of unwanted data collection and disclosure, its citizens need the power to make
`
`decisions about the fate of their unique biometric identifiers and information. Defendant
`
`Microsoft’s actions – including but not limited to their failure to provide the requisite notice and
`
`obtain the requisite consent – robbed Plaintiffs and Class Members of that power.
`
`80. Moreover, as a result of Defendant Microsoft’s unlawful conduct, Plaintiffs’ and
`
`Class Members’ biometric identifiers and information are no longer under their control and are
`
`CLASS ACTION COMPLAINT - 15
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`Case 2:20-cv-01082 Document 1 Filed 07/14/20 Page 16 of 26
`
`
`
`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
`
`
`
`available to a potentially unlimited range of unknown individuals for whatever uses they please.
`
`These injuries, which are imminent and clearly impending, are in addition to the injuries
`
`Plaintiffs and Class Members have already sustained as a result of Defendant’s actions.
`
`81.
`
`As a result of Defendant Microsoft’s misconduct, Plaintiffs and Class Members
`
`have no recourse for the fact that their biologically unique information has been compromised.
`
`82. Moreover, as a result of Defendant Microsoft’s misconduct, Plaintiffs and Class
`
`Members are likely to withdraw from biometric-facilitated transactions and other facially-
`
`mediated electronic participation.
`
`CLASS ACTION ALLEGATIONS
`
`83.
`
`Plaintiffs bring this action on behalf of themselves and as a class action under
`
`Federal Rule of Civil Procedure 23, seeking damages and equitable relief on behalf of the
`
`following Class for which Plaintiffs seek certification: All Illinois residents whose faces appear
`
`in the Diversity in Faces Dataset obtained by Defendant Microsoft.
`
`84.
`
`Excluded from the Class are: (a) Defendant Microsoft; (b) any parent, affiliate or
`
`subsidiary of Defendant Microsoft; (c) any entity in which Defendant Microsoft has a
`
`controlling interest; (d) any of Defendant Microsoft’s officers or directors; or (e) any successor
`
`or assign of Defendant Microsoft. Also excluded are any judge or court personnel assigned to
`
`this case and members of their immediate families.
`
`85.
`
`Plaintiffs reserve the right to amend or modify the class definitions with greater
`
`specificity or division after having had an opportunity to conduct discovery.
`
`86.
`
`Numerosity. While the exact number of Class Members is not known at this
`
`time, Defendant Microsoft obtained
`
`the biometric
`
`identifiers and
`
`information from
`
`approximately one million images of faces, and Plaintiffs estimate the total number of Class
`
`CLASS ACTION COMPLAINT - 16
`
`
`
`
`
`LOEVY & LOEVY
`100 S. KING ST., #100-748
`SEATTLE, WASHINGTON 98104
`T: 312-243-5900, FAX: 312-243-5092
`
`
`
`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-01082 Document 1 Filed 07/14/20 Page 17 of 26
`
`
`
`Members t