Turner v. Microsoft Corporation

CourtDistrict Court, N.D. California
DecidedMay 28, 2024
Docket4:22-cv-05827
StatusUnknown

This text of Turner v. Microsoft Corporation (Turner v. Microsoft Corporation) is published on Counsel Stack Legal Research, covering District Court, N.D. California primary law. Counsel Stack provides free access to over 12 million legal documents including statutes, case law, regulations, and constitutions.

Bluebook
Turner v. Microsoft Corporation, (N.D. Cal. 2024).

Opinion

1 2 3 4 UNITED STATES DISTRICT COURT 5 NORTHERN DISTRICT OF CALIFORNIA 6 7 DANA TURNER, et al., Case No. 22-cv-05827-DMR

8 Plaintiffs, ORDER 9 v. Re: Dkt. No. 49 10 NUANCE COMMUNICATIONS, INC., 11 Defendant.

12 This action began with two separately-filed putative class actions in which three named 13 plaintiffs alleged that Defendant Nuance Communications, Inc. (“Nuance”) violated the California 14 Invasion of Privacy Act (“CIPA”). The court consolidated the matters on May 31, 2023 and 15 Plaintiffs subsequently filed a consolidated amended class action complaint. [Docket No. 34 16 (“CFAC”).] Nuance moves to dismiss the CFAC for failure to state a claim under Federal Rule of 17 Civil Procedure 12(b)(6). [Docket No. 49 (“Mot.”).] The court held a hearing on April 11, 2024. 18 For the following reasons, the motion is granted in part and denied in part. 19 I. FACTUAL BACKGROUND 20 Plaintiffs make the following allegations in the CFAC, all of which are taken as true for purposes of the motion to dismiss.1 Nuance “markets and provides an artificial intelligence 21 software-as-a-service that allows businesses to authenticate their customers’ identities with their 22 voice,” called “Gatekeeper.” CFAC ¶ 1. Gatekeeper is a voice recognition software which 23 “records a consumer’s voice, creates a biometric voice print of the caller, and then examines that 24 voiceprint to determine whether the caller is a potential fraudster in future calls, and—above all— 25 26 1 When reviewing a motion to dismiss for failure to state a claim, the court must “accept as true all 27 of the factual allegations contained in the complaint.” Erickson v. Pardus, 551 U.S. 89, 94 (2007) 1 to determine whether the caller is telling the truth when identifying themselves and attempting to 2 gain access to their customer account.” Id. at ¶ 4. Businesses can license Gatekeeper and 3 integrate it into their call centers to verify callers’ identities. Id. at ¶¶ 5, 27. 4 Nuance’s marketing describes its Gatekeeper product as an “AI Risk Engine” which uses 5 “[d]eep neural networks” to “analyze biometric, non-biometric, and other available data to make 6 intelligent authentication and fraud risk assessments.” Id. at ¶ 31. Plaintiffs allege that Nuance 7 “analyzes ‘more than 1,000 characteristics’ of a person’s unique voice, such as the sound of a 8 person’s voice, how they talk, what they say, their pattern of speech, word choice, grammar, and syntax.” Id. at ¶ 2. Nuance can “authenticat[e] individuals by the way they talk . . . and flag[] 9 potential bad actors in as quick as half a second.” Id. at ¶ 28. Nuance’s technology works by 10 creating a voice print for each consumer derived from a recording and examination of the 11 consumer’s voice, enrolling those voice prints into a database, and then comparing the voice 12 characteristics of later callers against its saved voice prints. Id. at ¶ 29. A voice print serves as an 13 audible “fingerprint” which can “directly identify an individual.” Id. at ¶ 22. Plaintiffs allege that, 14 by “segment[ing]” and “cluster[ing]” voices and metadata from prior callers, “Nuance turns 15 individuals’ personal biometric data into a product offering for corporate clients.” Id. at ¶ 30. 16 Plaintiffs assert that Nuance has “amassed a massive database” of individuals’ voice prints and 17 retains that biometric data for “a significant period of time.” Id. at ¶¶ 32-33. 18 Plaintiffs claim that Nuance creates voice prints without the consumer’s “knowledge or 19 express written consent.” Id. at ¶ 37. Nuance’s software “seamlessly incorporates into its 20 customers’ call centers, without adequate notice (or any at all) that Nuance . . . is even involved in 21 the call.” Id. Plaintiffs allege that Nuance “listens to the consumer’s voice quietly in the 22 background of a call, and in such a way that consumers will likely be entirely unaware they are 23 unknowingly interacting with a third-party company.” Id. at ¶ 3. 24 According to Plaintiffs, Nuance’s voice recognition service exposes customers to 25 significant risks. For example, Plaintiffs state that voices are highly personal and can reveal 26 sensitive information about an individual’s mental state and behaviors. Id. at ¶ 6. They also aver 27 that artificial intelligence is “extremely susceptible to racial and gender bias,” and that voice 1 examination technology has been known to mistakenly classify legitimate customers as fraudsters 2 based on their voice, locking them out of their accounts and finances. Id. at ¶¶ 34-35. In addition, 3 they allege that Nuance’s system can be “easily fooled” by fraudsters, such as by using a free 4 online AI voice cloning service with an audio recording of a person’s voice. Id. at ¶ 36. 5 Plaintiff Turner called Chase’s customer support call center on numerous occasions, 6 including most recently in October 2022. Id. at ¶ 44. Turner alleges that she reasonably expected 7 her conversation with Chase to be confidential because Chase was a banking entity, “which 8 naturally involves the discussion of confidential information,” and because Turner spoke to Chase on her personal telephone and not in the direct presence of others. Id. at ¶ 45. During the call 9 with Chase, Turner was asked to make various “yes” or “no” statements in order to respond to 10 questions Chase asked her, or to otherwise provide additional information to Chase. Id. at ¶ 46. 11 Unbeknownst to Turner, her call was recorded by Nuance’s technology, which created a voice 12 print for Turner and automatically enrolled her voice print in Nuance’s biometric voice print 13 database. Id. at ¶¶ 47-48. Turner asserts that she did not give her consent, written or otherwise, to 14 “allow Nuance to wiretap her confidential communications with Chase.” Id. at ¶ 50-51. 15 Plaintiffs Smith and Youshei allege similar facts. They both called Chase’s customer 16 support call center multiple times during the past three years. Id. at ¶¶ 54, 63. They allege that 17 they reasonably expected their conversations with Chase to be confidential. Id. at ¶¶ 55, 64. 18 However, their calls were recorded by Gatekeeper, which created a voice print for each Plaintiff 19 and automatically enrolled the voice prints in Nuance’s biometric voice print database. Id. at ¶¶ 20 57, 66. Neither Nuance nor Chase disclosed to Plaintiffs that their voices were being recorded or 21 analyzed by Nuance to make a voice print, nor that their voice prints were being enrolled in 22 Nuance’s voice print database. Id. at ¶¶ 59, 68. Plaintiffs Smith and Youshei allege that they did 23 not consent to Nuance “collect[ing] [their] voice print[s] and to examine, record, wiretap, or 24 analyze [their] voice[s] for any purpose whatsoever.” Id. at ¶¶ 60, 69. 25 Plaintiffs assert that Nuance violated Cal. Penal Code §§ 631(a), 632(a), and 637.3. They 26 seek to represent a statewide class of similarly situated individuals, defined as: “All residents of 27 the State of California who had their voice prints, voice stress patterns, or other elements of their 1 conversation recorded and examined by Nuance without first obtaining prior written consent.” 2 CFAC ¶ 72. Plaintiffs seek declarative relief, injunctive relief, and damages. Id. at 10. 3 II. PROCEDURAL BACKGROUND 4 Turner brought a putative class action on October 6, 2022. [Docket No. 1.] Turner filed a 5 first amended class action complaint on January 9, 2023. [Docket No. 19.] Smith and Youshei 6 also brought a putative class action, which the court related to Turner’s case on March 30, 2023. 7 [Docket No. 27.] The cases were subsequently consolidated on May 31, 2023. [Docket No. 33.] 8 Plaintiffs filed the CFAC on June 14, 2023. Nuance now moves to dismiss the CFAC. 9 III. LEGAL STANDARD 10 A motion to dismiss under Rule 12(b)(6) tests the legal sufficiency of the claims alleged in 11 the complaint. See Parks Sch.

Free access — add to your briefcase to read the full text and ask questions with AI

Related

Erickson v. Pardus
551 U.S. 89 (Supreme Court, 2007)
Bell Atlantic Corp. v. Twombly
550 U.S. 544 (Supreme Court, 2007)
Shroyer v. New Cingular Wireless Services, Inc.
622 F.3d 1035 (Ninth Circuit, 2010)
Claire A. Straughn v. Delta Air Lines, Inc.
250 F.3d 23 (First Circuit, 2001)
Lee v. City Of Los Angeles
250 F.3d 668 (Ninth Circuit, 2001)
Ribas v. Clark
696 P.2d 637 (California Supreme Court, 1985)
United States v. Reed
575 F.3d 900 (Ninth Circuit, 2009)
Tavernetti v. Superior Court
583 P.2d 737 (California Supreme Court, 1978)
Coto Settlement v. Eisenberg
593 F.3d 1031 (Ninth Circuit, 2010)
Viviano v. Travelers Insurance
533 F. Supp. 1 (E.D. Louisiana, 1981)
Rogers v. Ulrich
52 Cal. App. 3d 894 (California Court of Appeal, 1975)
Parker v. Wakelin
887 F. Supp. 14 (D. Maine, 1995)
Flanagan v. Flanagan
41 P.3d 575 (California Supreme Court, 2002)
Kearney v. Salomon Smith Barney, Inc.
137 P.3d 914 (California Supreme Court, 2006)
Nancy Graf v. Zynga Game Network, Inc.
750 F.3d 1098 (Ninth Circuit, 2014)

Cite This Page — Counsel Stack

Bluebook (online)
Turner v. Microsoft Corporation, Counsel Stack Legal Research, https://law.counselstack.com/opinion/turner-v-microsoft-corporation-cand-2024.