In re Facebook Biometric Information Privacy Litigation

ORDER RE MOTION TO DISMISS AND SUMMARY JUDGMENT

JAMES DONATO, United States District Judge

In this putative class action under the Illinois Biometric Information Privacy Act, 740 Ill. Comp. Stat. 14/1 et seq. (“BIPA”), the named plaintiffs allege that defendant Facebook, Inc. (“Facebook”) unlawfully collected and stored biometric data derived from their faces. Although the case is brought by Illinois residents under Illinois law, it is before this Court because the parties agreed to transfer it here from the United States District Court for the Northern District of Illinois.

In the motions before the Court, Face-book argues that plaintiffs have failed to state a claim under BIPA and that a California choice-of-law provision in its user agreement precludes suing on an Illinois statute. Plaintiffs say the BIPA allegations do state a claim and that they never agreed to a choice of California law. Even if they had, plaintiffs contend Illinois law applies under traditional choice-of-law rules. After briefing and an evidentiary hearing on disputed fact issues underlying choice of law, the Court finds that Illinois law applies and that plaintiffs have stated a claim under BIPA.

BACKGROUND

As alleged in the complaint, Facebook “operates the largest social network in the world, with over one billion active users.” Dkt. No. 40 ¶ 1. The three named plaintiffs, Nimesh Patel, Adam Pezen and Carlo Licata, are Facebook users who “use its platform to, among other things, upload and share photographs with friends and relatives.” Id. ¶¶ 2, 7-9.

This case arises out of Facebook’s “Tag Suggestions” program, which was launched in 2010. Id. ¶ 3. A “tag” on Face-book is when a user identifies by name other Facebook users and non-users who appear in the photographs that have been uploaded to Facebook. Id. ¶ 2. “Tag Suggestions” is intended to encourage more tagging on Facebook. Id. ¶ 3. The program functions by scanning uploaded photographs “and then identifying faces appearing in those photographs.” Id. If the program “recognizes and identifies one of the faces appearing in [a] photograph, Face-book will suggest that individual’s name or automatically tag them.” Id. In effect, the program puts names on the faces in photos and prompts users to tag those individuals.

To make the tagging suggestions, Face-book utilizes “state-of-the-art facial recognition technology” to extract biometric identifiers from the profusion of photographs users upload. Id. ¶¶ 4,22. Facebook creates and-' stores digital representations (known as “templates”) of people’s faces based on the geometric relationship of facial features unique to each individual, “like the distance between [a person’s] eyes, nose and ears.” Id. ¶ 23.

■Plaintiffs allege that Facebook amassed users’ biometric data secretly and without consent.' Specifically, they allege that the Tag Suggestions program violated BIPA because Facebook did not: “[1] properly inform plaintiffs or the class in writing that their biometric identifiers (face geometry) were being generated, collected or stored; [2] properly inform plaintiffs or the class in writing of the specific purpose-and length of time for which their biometric identifiers were being collected, stored, and used; [3] provide a publicly available retention schedule and guidelines for permanently destroying the biometric identifiers of plaintiffs and the class (who do not opt-out of ‘Tag Suggestions’); and [4] receive a written release from plaintiffs or the class to collect, capture, or otherwise obtain their biometric identifiers.” Id. ¶ 5. Plaintiffs seek declaratory and injunctive relief, and statutory damages. Id. ¶ 6.

This case was previously pending as three separate cases (one for each of the three named plaintiffs) in the federal district court for the Northern District of Illinois. Dkt. No. 1; The parties stipulated to transfer them to this Court. Dkt. No. 29. Once here, the Court consolidated the three cases into a single action. Dkt. No. 44. Plaintiffs’ consolidated class action complaint, Dkt. No. 40, is the operative complaint.

Facebook filed a motion to dismiss the consolidated complaint under Rule 12(b)(6). Dkt. No. 69. The motion made two arguments: (1) “plaintiffs cannot pursue a claim under the Illinois BIPA because they agreed that California law governs their disputes with Facebook”; and (2) “the Illinois BIPA does not apply to Tag Suggestions.” Id. at, 6,10.

'. Plaintiffs denied that they agreed to Fa-.cebook’s user, agreement, including the choice-of-law provision, and raised fact disputes, that could not be resolved within the confines of a Rule 12(b)(6) motion. The Court converted this portion of defendant’s motion to dismiss into a summary judgment proceeding under Rule 56, and set an evidentiary hearing on the contract formation dispute for the choice-of-law provision. Dkt. No. 85 at 20:14-23, 24:3-11. Defendant’s second argument for dismissal— that plaintiffs had failed to state a claim under BIPA — was taken under submission pending resolution of the choice-of-law question.

After additional briefing from both sides, Dkt. Nos. 96, 97-3, the Court held an evidentiary and summary judgment hearing. As the proponent of the choice-of-law provision, Facebook called two live witnesses: Joachim De Lombaert, a Face-book engineering manager, and Mark Pike, a Facebook privacy program manager. Dkt. Nos. 96, 109. Plaintiffs cross-examined these witnesses and presented portions of,.each of the three plaintiffs’ videotaped depositions, but did not call any live witnesses of their own. After the evidentiary hearing, the Court took arguments from counsel on the summary judgment issues: whether a contract had been formed -on choice of law, and if so, whether it should be enforced to bar plaintiffs from asserting claims under Illinois law. Dkt. No. 109.

This order resolves the motion to dismiss and the motion for summary judgment. All previously unaddressed evi-dentiary objections are overruled and .plaintiffs’ request to defer summary judgment under Rule 56(d) is denied.

DISCUSSION

I. SUMMARY JUDGMENT

A. CHOICE-OF-LAW FACTFINDING

Before getting to the findings of fact from the evidentiary hearing, the Court addresses the question of whether it should be making any findings at all. Plaintiffs object that, “[t]o the extent the evidentiary hearing on the issue of assent would involve weighing evidence, assessing credibility of live testimony, and resolving disputed issues of fact, that would invade the province of the jury.” Dkt. No. 97-3 at 1-2. This objection is cursory — plaintiffs devote less than one page to it in their brief — and unpersuasive.

It is certainly true, and not disputed by Facebook, that plaintiffs have a right to a jury trial in this case under the Seventh Amendment to the United States Constitution. But the attachment of a jury right to the case as a whole does not mean that each and every issue in the case “is itself necessarily a jury issue.” Markman v. Westview Instruments, Inc., 517 U.S. 370, 378, 116 S.Ct. 1384, 134 L.Ed.2d 577 (1996). In appropriate circumstances, the Court may decide fact disputes raised on the way to trial. For example, our Circuit has established that “a judge rather than a jury decides disputed factual questions relevant to jurisdiction and venue.” Albino v. Baca, 747 F.3d 1162, 1170-71 (9th Cir.2014). And the Supreme Court held in Markman that, in a patent infringement case otherwise subject to trial by jury, the issue of “the construction of a patent, including terms of art within its claim, is exclusively within the province of the court.” 517 U.S. at 372, 116 S.Ct. 1384. That is so even though claim construction has “evidentiary underpinnings” and may involve “credibility judgments” about witnesses. Id. at 389-90, 116 S.Ct. 1384; see also Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., — U.S. -, 135 S.Ct. 831, 836-37, — L.Ed.2d - (2015) (applying clearly erroneous standard in FRCP 52(a) to “district court’s resolution of subsidiary factual matters made in the course of’ claim construction).

When a question arises of whether the judge or a jury should decide a fact dispute, Markman directs the courts to turn to history in the first instance for an answer. The right of trial by jury preserved in the Seventh Amendment “ ‘is the right which existed under the English common law when the Amendment was adopted.’ ” Markman, 517 U.S. at 376, 116 S.Ct. 1384 (quoting Baltimore & Carolina Line, Inc. v. Redman, 295 U.S. 654, 657, 55 S.Ct. 890, 79 L.Ed. 1636 (1935)). This “historical test” is easy to answer when there is “clear historical evidence that the very subsidiary question was so regarded under the English practice of leaving the issue for a jury.” Id. at 376-77, 116 S.Ct. 1384. But if the past does not speak clearly, “we are forced to make a judgment about the scope of the Seventh Amendment guarantee without the benefit of any foolproof test.” Id.

The historical inquiry here is unhelpful. Plaintiffs have not proffered any historical evidence in support of their objection, either because there isn’t any or they didn’t take the time to look — the Court cannot tell. They also do not cite to any guidance in the form of controlling case law. Instead, they have pointed only to two out-of-district cases, Dkt. No. 97-3 at 2, one of which offers the observation that the “Circuits appear to be in dispute as to whether the court should resolve factual issues relevant to the choice of law.” Mattel, Inc. v. MCA Entm’t, Inc., 782 F.Supp.2d 911, 977 n. 24 (C.D.Cal.2011).

The law is clearer than that characterization suggests. The Fifth and Seventh Circuits have expressly approved judicial resolution of fact disputes raised in a choice-of-law detemination. See Vaz Borralho v. Keydril Co., 696 F.2d 379, 386 (5th Cir.1983); Nautilus Ins. Co. v. Reuter, 537 F.3d 733, 742-43 (7th Cir.2008) (“[W]e are of the opinion that the district court must resolve factual disputes that bear on the choice-of-law determination.”). As the Seventh Circuit noted, similar matters like subject matter and personal jurisdiction, venue and abstention are decided by the judge “even if there are contestable factual questions bearing on the decision.” Nautilus, 537 F.3d at 743 (internal citation omitted).

Why should resolution of contestable facts for a choice-of-law determination be different? Plaintiffs do not say and there are substantial reasons not to treat it so. While the Ninth Circuit does' not appear to have directly addressed this issue, the Court finds that it would embrace the positions of the Fifth and Seventh Circuits. The Ninth Circuit has clearly accepted that the judge, not a jury, typically resolves the question of which forum’s law ought to be applied in a case. See, e.g., Gen. Signal Corp. v. MCI Telecomm. Corp., 66 F.3d 1500, 1505-06 (9th Cir.1995) (holding on review of district court’s choice-of-law decision that the circuit court “reviews decisions concerning the appropriate choice of law de novo” and making finding “that the JDA choice of law provision applies to disputes arising out of the Letter Agreement.”). In our Circuit, this determination is frequently made by the district judge on a summary judgment motion. See, e.g., In re TFT-LCD (Flat Panel) Antitrust Litig., No. 10-05625 SI, 2013 WL 6327490 (N.D.Cal. Dec. 3, 2013) (granting in part and denying in part defendant’s motion for partial summary judgment on choice-of-law grounds).

These factors point strongly to the conclusion that the Ninth Circuit would agree that the judge should resolve fact disputes subsumed in a choice-of-law determination. This conclusion is consistent with Supreme Court precedent. See Markman, 517 U.S. at 390, 116 S.Ct. 1384 (claim construction given to judge “notwithstanding its eviden-tiary underpinnings.”). And it is perfectly sound when viewed from the perspective of good case management. When “history and precedent provide no clear answers, functional considerations also play their part in the choice between judge and jury” in resolving fact disputes. Id. As other courts have rightly observed, the functional and practical results of assigning choice-of-law fact determinations to a jury are problematic, to say the least. See Nautilus, 537 F.3d at 743 (“[I]t would make little sense to let a jury decide which facts are true and then to say that there was never a dispute to begin with.”) (internal citation omitted); Toll v. Tannenbaum, 982 F.Supp.2d 541, 551-52 (E.D.Penn.2013) (presenting “choice-of-law factual issues to a jury would be impractical and confusing”).

The reasons why that would be a bad practice are self-evident. The litigants and the fair and efficient administration of justice would suffer immensely from slogging through all the pretrial activities of discovery, class certification, and dispositive motions, and then a full trial, without knowing which law governs the case. The consequences of doubled or trebled litigation costs, destabilizing uncertainty about dispute outcomes, and overall case management chaos are too plain to be debated. And the Court can only imagine with apprehension what jury instructions and verdict forms would look like in a case that required the jury to first pick the governing law.

Consequently, the best approach is for the. Court to resolve fact disputes subsumed in deciding choice of law. To be sure, this result might not apply in every case. In a rare situation, it is possible that a choice-of-law fact dispute is so bound up in the substantive claims that the court cannot decide it without compromising the constitutional guarantee of a jury resolution. Plaintiffs cite two cases where that rarity, was discussed, and use them to argue that choice-of-law fact disputes should always be reserved for a jury. Dkt. 97-3 at 2. But those cases turned on unusual circumstances not present here, and neither stands for the broad rule that plaintiffs urge. Marra v. Bushee, 447 F.2d 1282 (2d Cir.1971), for example, involved a loss of consortium claim arising out of an alleged marital affair in two different states. Choice of law depended on deciding where the affair primarily occurred, but since that issue — whether the defendant had in fact been involved in an affair and caused loss of consortium — was effectively the ultimate dispute in the case, the court concluded that a jury rather than the judge should determine “the territorial location of [defendant’s] alluring conduct” and consequent choice of law. Id. at 1284. Similarly, in Mattel, 782 F.Supp.2d at 976-77, the parties in a trade secrets case disputed where an alleged misappropriation of documents happened, which would affect choice of law. Resolution of that issue would again go directly to the merits of the claim — whether misappropriation had occurred — and the court reserved that “foundational dispute” for the jury although it would, as in Marra, also decide the choice-of-law question,

This case is unlike Marra and Mattel because the resolution of the fact issues subsumed in choice of law will not usurp the right to a jury determination of the merits of plaintiffs’ claims. In those cases, the merits of the substantive claims were inextricably intertwined with the choice-of-law determination. The proper law to apply could not be determined without also deciding whether the defendants were liable for misappropriation or loss of consortium — questions indisputably subject to jury determination. That overlap of choice-of-law facts and the merits does not exist in this case. The question of whether plaintiffs agreed to Facebook’s user agreement does not require resolution of the BIPA claims and has nothing at all to do with biometrics. Because the merits of plaintiffs’ substantive claims are wholly irrelevant to deciding the choice-of-law issue, the Court can resolve the fact disputes subsumed in the choice-of-law decision in this case while fully preserving plaintiffs’ right to a jury’s resolution of the ultimate dispute.

Plaintiffs’ objection to the evidentiary hearing is misplaced. The Court will resolve the factual dispute of whether or not plaintiffs consented to a California choice-of-law provision as argued by Facebook.

B. FINDINGS OF FACT

The Court finds that the following facts were established at the evidentiary hearing. Dkt. No. 109. The parties agree that the applicable standard of proof is a preponderance of the evidence. See, e.g., Dkt. No. 112 at 113:23-114:5.

1. Initial Registration

Plaintiff Adam Pezen

1. For Pezen to complete the registration process for Facebook on August 22, 2005, he would have had to click a box next to the words, “I have read and understood the Terms of Use, and I agree to them.” Dkt. No. 112 at 20:17-23:21. In that sentence, the words “Terms of Use” were highlighted, and clicking on those words would have taken Pezen to a separate page on which he would have seen the current terms of use as of that date. Id.

2. In addition to clicking that check box, Pezen would have had to provide his name, status, email and his password of choice. He then would have had to complete the process by clicking a button that said, “Register Now!”

3. It was not possible to sign up for Facebook using a mobile phone in 2005. Signing up via computer was the only option, Id. at 9:12-14. This did not change until early 2008, Id. at 9:18-19.

Plaintiff Nimesh Patel

4. At his deposition, Patel testified that his “best recollection” was that he registered for Facebook “on a computer,” rather than on his smartphone. Dkt. No. 112 at 48:6-8.

5. A new user joining Facebook via computer on February 11, 2008, as Patel did, would have seen one of two screens. One version asked for more information than the other. The shorter version sought the new user’s full name, birthday, email address and a new password. The other version additionally asked whether the user was “in college/graduate school, at a company, in high school, [or] none of the above,” and for the user’s high school name and graduation year. The two different versions were presented to the public at random, and most users would have seen the longer version of the sign-up screen. Id. at 25:6-16:10,

6. In both versions, the user signing up via computer would have had to click a box next to the words, “I have read and agree to the Terms of Use and Privacy Policy.” In that sentence, the words “Terms of Use” and “Privacy Policy” were highlighted. Both sets of highlighted words constituted links, and by clicking on them, users could actually read the Terms of Use and Privacy Policy. Id. at 29:4-14. This was true in both versions of the sign-up page users would have seen on'February 11, 2008.

7. In either version, the user would have had to complete the process by clicking a button that said, “Sign Up.”

8. A new user joining via a mobile device or phone would not have been presented with a check box next to the words, “I have read and agree — ” Instead, the user would have seen a sentence that said, “By submitting this information, I acknowledge that I have read and agree[ ] to the Terms of Use and Privacy Policy,” “followed by a form submission button,” ie,, a button that said, “Sign Up,” immediately below. Id. at 27:24-28:13, 30:7-23. Plaintiff Carlo Licata

9. At his deposition, Licata testified that he did “not recall what phone or computer tliat [he] used” to sign up for Facebook. Dkt. No. 112 at 49:12-15.

10. A new user joining Facebook via computer on November 13, 2009, which is when Licata signed up for his Facebook account, would have had to work through two screens as part of a two-step process. Id. at 30:24-32:10. On the first screen, the user would have had to provide his first and last names, email address, a new password, and indicate his gender and birthday. Clicking the “Sign Up” button on the bottom of the screen would then have taken him to the second screen. In that screen, the user would have had to complete a “CAPTCHA” security check, typically typing in words that corresponded to the image displayed. Id. at 32:13-33:9. After typing those words into the box provided, the user would have had to click another “Sign Up” button to complete the process. Immediately below the second “Sign Up” button was this sentence: “By clicking Sign Up, you are indicating that you have read and agree to the Terms of Use and Privacy Policy.” The words “Terms of Use” and “Privacy Policy” were highlighted and represented links to other pages on which those documents could actually be found. Id. at 33:11-19, 36:2-24.

11. As in 2008, a new user joining via a mobile device or phone in November 2009 would not have been presented with a check box next to the words, “I have read and agree.... ” Instead, the user would have seen a sentence that said, “By submitting this information, I acknowledge that I have read and agree[ ] to the Terms of Use and Privacy Policy,” “followed by a form submission button,” ie., a button that said, “Sign Up,” immediately below. Id. at 27:24-28:13, 30:7-23.

2. Current Terms of Use

12. The current Terms of Use are those that were last revised on January 30, 2015. Dkt. No. 112 at 86:15-18. Paragraph 15.1 of the Terms states, “The laws of the State of California will govern this Statement, as well as any claim that might arise between you and us, without regard to conflict of law provisions.” Id. at 100:25-101:25.

13. Facebook emailed all users with registered email addresses about this update. Id. at 87:14-15, 88:14-15. The email was entitled, “We’re updating our terms and policies and introducing Privacy Basics,” and the email contained hyperlinks to the new Terms of Use. Id. at 94:15-97:24.

14. Facebook also provided “jewel notifications” about the update. On the “notification tray at the top of the site” which “turns red and maybe show[s] you a number to show you how many notifications you have,” Facebook “changed that badge to indicate that you had a new message.” “[I]f you clicked on it, it would let you know that.. .there [were] new Terms proposed and getting updated.” Id. at 89:6-15. So when a user logged on to Facebook.com and was on his or her “news feed,” the user “would see the jewel notification,” and “that persisted a number of times in case [the user] didn’t visit.” Id. at 90:9-11.

15. In addition to these more individualized forms of notice, Facebook also “published updates from the Facebook’s governance page so that people who are logged into Facebook could see updates on the page, with relevant information about the update”; it “launched something called a Privacy Basi[c]s Center, which explained in a little bit more detail some of the updates in the terms”; it utilized the “Fa-cebook News Room” to publish short news items about updates including to the terms; and for those who “were viewing the terms page, [Facebook] also displayed something that’s called a rooster or a banner that was displayed at the top.” Id. at 87:18-88:5.

16.Users were not required to click anything or otherwise take any affirmative steps to be subjected to the current Terms of Use. Facebook took users' continued use as assent to the terms. Id. at 91:11-20.

C. CONCLUSIONS OF LAW

1. A Choice-of-Law Agreement Was Formed

“‘While new commerce on the Internet has exposed courts to many new situations, it has not fundamentally changed the principles of contract.’” Nguyen v. Barnes & Noble Inc., 763 F.3d 1171, 1175 (9th Cir.2014) (quoting Register.com, Inc. v. Verio, Inc., 356 F.3d 393, 403 (2d Cir.2004)); see also One Beacon Ins. Co. v. Crowley Marine Servs., Inc., 648 F.3d 258, 268 (5th Cir.2011) (same); Treiber & Straub, Inc. v. U.P.S., Inc., 474 F.3d 379, 385 (7th Cir.2007) (same). The touchstone for all contracts, whether Internet digital or old school paper, is “ ‘mutual manifestation of assent, whether by written or spoken word or by conduct.’” Nguyen, 763 F.3d at 1175 (quoting Specht v. Netscape Commc’ns Corp., 306 F.3d 17, 29 (2d Cir.2002)). To determine whether a binding contract has been formed, the dispositive questions are (1) did the offeror “provide reasonable notice” of the proposed terms, and (2) did the offeree “unambiguously manifest assent” to them? Nguyen, 763 F.3d at 1173.

To answer those questions for contracts formed on the Internet, our Circuit has settled on an analytical framework that puts “clickwrap” agreements at one end of the enforceability spectrum and “browsewrap” agreements at the other. On the issue of the notice provided by the offeror, a clickwrap agreement differs from a browsewrap agreement in this way: for clickwrap agreements, the user is “presented with a list of terms and conditions of use” before being asked to agree to them. Nguyen, 763 F.3d at 1176; see also Fteja v. Facebook, Inc., 841 F.Supp.2d 829, 837-38 (S.D.N.Y.2012) (describing “pure-form clickwrap agreement” as one in which the mechanism of the website “forces the user to actually examine the terms before assenting.”). For browsewrap agreements, the “website’s terms and conditions of use are generally posted on the website via a hyperlink at the bottom of the screen.” Nguyen, 763 F.3d at 1175-76; see also Fteja, 841 F.Supp.2d at 835-36 (“browse-wrap agreement usually involves... Terms of Use not listed on the site itself but available only by clicking on a hyperlink.”).

The two “wraps” also differ in the way the offeree manifests assent. For clickwrap agreements, users are “required to click on an T agree’ box”; they must expressly manifest assent to the terms and conditions. Nguyen, 763 F.3d at 1175-76. For browsewrap agreements, the user “gives his assent simply by using the website.” Id. at 1176 (quotation omitted). “Indeed, ‘in a pure-form browsewrap agreement, the website will contain a notice that — by merely using the services of, obtaining information from, or initiating applications within the website — the user is agreeing to and is bound by the site’s terms of service.’ ” Id. (quoting Fteja, 841 F.Supp.2d at 837).

The features distinguishing the wraps in practice may be summarized this way:

[[Image here]]

Although there is no per se rule of validity or invalidity on either end, our Circuit has recognized that the closer digital agreements are to the clickwrap end of the spectrum, the more often they have been upheld as valid and enforceable. See Nguyen, 763 F.3d at 1176 (“[cjourts have... been more willing to find the requisite notice for constructive assent where the browsewrap agreement resembles a clickwrap agreement”); see also Savetsky v. Pre-Paid Legal Servs., Inc., No. 14-03514 SC, 2015 WL 604767, at *3 (N.D.Cal. Feb. 12, 2015) (“courts generally find that clickwrap agreements are enforceable”). Browsewrap agreements, on the other hand, have been viewed with skepticism. See Nguyen, 763 F.3d at 1178-79 (observing the “courts’ traditional reluctance to enforce browsewrap agreements against individual consumers,” and declining to find contract had formed where “a website makes its terms of use available via a conspicuous hyperlink on every page of the website but otherwise provides no notice to users nor prompts them to take any affirmative action to demonstrate assent”).

Applying the Nguyen framework here, the evidence shows that the user agreements the plaintiffs encountered generally fell toward the browsewrap end of the spectrum. In each case, the terms of use were available via a hyperlink to a different page and not presented immediately to plaintiffs for review. But a typical clickwrap feature was also in place for two of the plaintiffs. Pezen and Patel had to click a box separately affirming that they had read and agreed to the Terms of Use. For Pezen, the .words next to the box were, “I have read and understood the Terms of Use, and I agree to them.” For Patel, they were, “I have read and agree to the Terms of Use and Privacy Policy.” In each case, these boxes were separate from, and in addition to, the “Register Now!” or “Sign Up” button that needed to be pi’essed to complete the entire process. On the question of computer vs. mobile phone, for Pezen, Facebook has shown that signing up by mobile phone was not even an option at the time he signed up. For Patel, too, Facebook has sufficiently shown at this stage that Patel is more likely to have signed up via computer.

Plaintiff Licata had a different and more questionable experience. Whether he signed up by computer or mobile device, he was not required to click a box specifically and separately manifesting his assent to the user agreement. Instead, he was asked only to click a “Sign Up” box with language under it that purported to put him on notice that clicking on it also constituted assent to the user agreement. If he signed up by computer, that language would have read, “By clicking Sign Up, you are indicating that you have read and agree to the Terms of Use and Privacy Policy,” with the terms of use presented by hyperlink only. The language would have been similar if he had signed up by mobile phone.

The procedure Licata encountered raises concerns about contract formation. The use of a single “Sign Up” click to activate an account and accept the terms of service presents a serious question of whether Fa-cebook provided reasonable notice of its agreement terms and whether the user truly manifested assent to them. If this Court were deciding those questions on a clean slate, it might find that users in Licata’s situation did not form a contract with Facebook. But our Circuit has indicated a tolerance for the single-click “Sign Up” and assent practice. In Nguyen, the Circuit cited with approval a decision from the Southern District of New York finding that this approach by Facebook was enough to create an enforceable agreement. Nguyen, 763 F.3d at 1176-77 (citing Fteja, 841 F.Supp.2d at 838-40). What appears to save the approach is that a user like Licata had to take some action — a click of a dual-purpose box — from which assent might be inferred. A contract was not foisted upon him simply by passively viewing a website. Nguyen advises that this difference is enough to form a contract.

The Court consequently finds plaintiff Licata assented to the user agreement, and the Court also finds the same to be true of plaintiffs Pezen and Patel, for whom the manifestation of assent was more clear. See also Crawford v. Beachbody, LLC, No. 14cv1583-GPC (KSC), 2014 WL 6606563, at *3 (S.D.Cal. Nov. 5, 2014) (observing that “[c]ourts have held that a modified or hybrid click-wrap/browsewrap agreement constitutes a binding contract where the user is provided with an opportunity to review the terms of service in the form of a hyperlink immediately under the T Accept’ button and clicks that button,” and enforcing such an agreement in that case); Zaltz v. JDATE, 952 F.Supp.2d 439 (E.D.N.Y.2013) (in case where terms and conditions appeared by hyperlink and user had to click on a separate box confirming that he or she has read and agreed to the terms, finding cir-cumstanees analogous to those in Fteja and upholding contract). Facebook has satisfactorily shown that it is more likely than not that all three plaintiffs assented to the user agreement when they signed-up for Facebook.

The Court also finds that Facebook' has shown by a preponderance of the evidence that all three plaintiffs agreed to the current user agreement. They were provided notice that the terms of the user agreement were changing through an email from Facebook sent directly to the email addresses each plaintiff had on file with Facebook. Each .plaintiff — -none of whom disputes remaining an active Facebook user to this day, see e.g., Dkt. No. 95-3 at 6 — would also have received a “jewel notification” on his individual Facebook news-feed. This individualized notice in combination with a user’s continued use is enough for notice and assent. Cf. Rodman v. Safeway Inc., Case No. 11-cv-03003-JST, 2015 WL 604985, at *10-11 (N.D.Cal. Feb. 12, 2015) (“[E]ven in light of their agreement to the Special Terms at the time of registration, customers’ assent to the revised Terms cannot be inferred from their continued use of Safeway.Com when they were never given notice that the Special Terms had been altered”; noting that Safeway could have, but did not, “send all existing Safeway.com customers an email in order to ensure that every consumer is aware of a change in the Special Terms prior to making a purchase”).

Plaintiffs do not dispute that the current user agreement contains a California choice-of-law provision. Because the user agreement is clearly contractual and was unambiguously presented that way, the choice-of-law provision did not need to be specially called out to users in' order to create a contract on that issue. All three plaintiffs initially accepted the user agreement that was then in force when each plaintiff created his Facebook account. Subsequently) plaintiffs were given adequate notice of the terms in the current user agreement, and the plaintiffs accepted and agreed to the current terms by continuing to use Facebook after receiving that notice. Plaintiffs are parties to a contractual California choice-of-law provision.

2. The Contractual Choice of Law Will Not Be Enforced

A) Governing Choice-of-Law Rules

The question now is whether the agreed-upon choice-of-law provision should be enforced. As an initial matter, the parties seemed to agree in their briefs that California choice-of-law rules apply. Mot. to Dismiss, Dkt. No. 69 at 6 (“[A] federal court sitting in diversity must look to the forum state’s choice-of-law rules.”); Pls’ Opp., Dkt. No. 73 at 3, n. 1 (assuming California choice-of-law rules apply without discussion). But at the hearing, Face-book urged the Court to look to “the way Illinois courts act when they are faced with a choice-of-law clause.” Dkt. No. 112 at 121: 8-10. This shift is potentially significant because California and Illinois choice-of-law rules are not identical.¹

Facebook’s reference to Illinois was error because California’s choice-of-law rules govern in this case. A “federal court sitting in diversity ordinarily must follow the choice-of-law rules of the State in which it sits.” Atl. Marine Const. Co. v. U.S. Dist. Court for W. Dist. of Texas, — U.S. -, 134 S.Ct. 568, 582, 187 L.Ed.2d 487 (2013) (citing Klaxon Co. v. Stentor Elec. Mfg. Co., 313 U.S. 487, 494-96, 61 S.Ct. 1020, 85 L.Ed. 1477 (1941)). “This applies to actions brought under the Class Action Fairness Act as well, since CAFA is based upon diversity jurisdiction.” In re NVIDIA GPU Litig., No. C 08-04312, 2009 WL 4020104, at *5 (N.D.Cal. Nov. 19, 2009). The fact that this case was transferred from the Northern District of Illinois is of no moment in deciding the applicable choice-of-law rules. The rule in Van Dusen v. Barrack, 376 U.S. 612, 638-39, 84 S.Ct. 805, 11 L.Ed.2d 945 (1964), that a transferee court should apply the law of the transferor court to avoid venue gamesmanship does not apply when, as here, the transfer is related to a venue agreement. Atl. Marine, 134 S.Ct. at 583. If that were not the case, another form of venue gamesmanship — trumping a forum-selection clause simply by filing outside the forum— would flow from Van Dusen. The Face-book user agreement provides for venue in this district and the parties stipulated to transfer here after Facebook moved to enforce the forum-selection clause. Dkt. No. 39 at 3. Consequently, the Court will follow the choice-of-law rules of its home state, California.

B) Enforceability of the California Choice of Law

In Washington Mutual Bank, FA v. Superior Court, 24 Cal.4th 906, 103 Cal.Rptr.2d 320, 15 P.3d 1071 (2001), the California Supreme Court set out the test to determine the enforceability of a contractual choice of law. The test follows Section 187 of the Restatement (Second) of Conflict of Laws and is premised on the “ ‘strong policy considerations favoring the enforcement of freely negotiated choice-of-law clauses.’ ” Id. at 917, 103 Cal.Rptr.2d 320, 15 P.3d 1071 (quoting Nedlloyd Lines B.V. v. Superior Court, 3 Cal.4th 459, 462, 11 Cal.Rptr.2d 330, 834 P.2d 1148 (1992)). “The parties’ choice generally will be enforced unless the other side can establish both that the chosen law is contrary to a fundamental policy” of the state law alternative to the contractual choice, and that the other state “has a materially greater interest in the determination of the matter.” Id.; see also Ruiz v. Affinity Logistics Corp., 667 F.3d 1318, 1323 (9th Cir.2012) (“California courts apply the parties’ choice of law unless the analytical approach articulated in § 187(2) of the Restatement (Second) of Conflict of Laws... dictates a different result.”).

While some of the Section 187 factors tilt in Facebook’s direction, this case presents strong reasons to depart from the parties’ contractual choice of law. As an initial matter, Facebook, as the advocate of the clause, “has met its burden of establishing that the various claims of putative class members fall within its scope.” Wash. Mut. Bank, 24 Cal.4th at 916, 103 Cal.Rptr.2d 320, 15 P.3d 1071. Facebook’s user agreement provides that “any claim that might arise between you and us” is subject to California law. Dkt. No. 70-1, Ex. A at 10. That is broad enough to encompass this dispute and plaintiffs do not seriously argue otherwise.

The next step “is to evaluate the clause’s enforceability pursuant to the analytical approach reflected in section 187” of the Restatement. Wash. Mut. Bank, 24 Cal.4th at 916, 103 Cal.Rptr.2d 320, 15 P.3d 1071. Under that approach, the first question is whether the chosen state has a substantial relationship to the parties or their transaction, or whether there is another reasonable basis for the parties’ choice of law. Id. This test is also readily met. Facebook has its principal executive offices and corporate headquarters in California, which satisfies this inquiry. See Peleg v. Neiman Marcus Grp., Inc., 204 Cal.App.4th 1425, 1447, 140 Cal.Rptr.3d 38 (2012) (substantial relationship and reasonable basis shown when corporation has its principal place of business and headquarters in choice of law state); Ruiz, 667 F.3d at 1323 (same).

While these initial factors count in favor of enforcing the parties’ choice of California law, the remaining tests weigh heavily against it. The final issues for analysis are “whether the chosen state’s law is contrary to a fundamental policy” of the alternative state’s law, and if so, whether that state “has a ‘materially greater interest than the chosen state in the determination of the particular issue’ ” presented in the case. Wash. Mut. Bank, 24 Cal.4th at 916, 103 Cal.Rptr.2d 320, 15 P.3d 1071 (quoting Restatement Section 187(2)) (emphasis in original). The parties agree that Illinois law applies if the California choice-of-law provision is ineffective, Dkt. No. 69 at 7-8; Dkt. No. 73 at 4-6, and so the specific issues here are whether the, California choice-of-law clause is, contrary to a fundamental policy of Illinois, and if so, whether Illinois has a greater interest in the determination of this case.

The answer to both questions is yes. There can be no reasonable doubt that the Illinois Biometric Information Privacy Act embodies a fundamental policy of the state of Illinois. “To be fundamental within the meaning of Restatement section 187, a policy must be a substantial one.” Brack v. Omni Loan Co., 164 Cal.App.4th 1312, 1323, 80 Cal.Rptr.3d 275 (2008). By its express terms, BIPA manifests Illinois’ substantial policy of protecting its citizens’ right to privacy in their personal biometric data. BIPA is premised on the Illinois legislature’s stated concerns about the use of new technology by “[m]ajor national corporations” to collect personal biometric data. 740 Ill. Comp. Stat. 14/5(b). The legislature expressly found that: (1) “Biometrics are unlike other unique identifiers. . .[and] 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” {id. at 14/5(c)); (2) “[t]he full ramifications of biometric technology are not fully known” {id. at 14/5(f)); and (3) “[t]he public welfare, security and safety will be served by regulating the collection, use, safeguarding, handling, storage, retention, and destruction of biometric identifiers and information” {id. at 14/5(g)). In connection with these findings, the Illinois legislature implemented a series of safeguards intended to protect the privacy of personal biometric data. Among other protections, BIPA requires written policies on biometric data retention and informed consent before obtaining or disclosing personal biometric data. Id. at 14/15(a), (b), (d). It bans the sale or trade of personal biometric information for profit. Id. at 14/15(c). And it provides for a private right of action to “[a]ny person aggrieved by a violation” of the statute. Id. at 14/20. Taking all of these factors as a whole, the plain language of BIPA indisputably evinces a fundamental privacy policy of Illinois.

It is equally undeniable that enforcing the contractual choice of California law would be contrary to this policy in the starkest way possible. Facebook tries to downplay the conflict as merely the loss of a claim. Dkt. No. 69 at 7-8. But if California law is applied, the Illinois policy of protecting its citizens’ privacy interests in their biometric data, especially in the context of dealing -with “major national corporations” like Facebook, would be written out of existence. That is the essence of a choice-of-law conflict. See, e.g., Ruiz, 667 F.3d at 1324 (declining to enforce Georgia choice of law that conflicted with “a fundamental California policy that seeks to protect its workers”). The conflict is all the more pronounced because California has no law or policy equivalent to BIPA. Unlike Illinois, California has not legislatively recognized a right to privacy in personal biometric data and has not implemented any specific protections for that right or afforded a private cause of action to enforce violations of it. See In re DirecTV Early Cancellation Litig., 738 F.Supp.2d 1062, 1088 (C.D.Cal.2010) (declining to enforce Florida, Illinois, New Jersey, Oregon, and Virginia choice-of-law provisions because the chosen states lacked consumer protection statutes comparable to California’s).

Illinois’ greater interest in the outcome of this BIPA dispute is also readily apparent. The fundamental question on this point is “which state, in the circumstances presented, will suffer greater impairment of its policies if the other state’s law is applied.” Bridge Fund Capital Corp. v. Fastbucks Franchise Corp., 622 F.3d 996, 1004 (9th Cir.2010) (citing Brack, 164 Cal.App.4th at 1329, 80 Cal.Rptr.3d 275). The answer here could not be clearer. Illinois will suffer a complete negation of its biometric privacy protections for its citizens if California law is applied. In, contrast, California law and policy will suffer little, if anything at all, if BIPA is applied. Facebook makes the implausible argument that California has the superior interest of needing to provide “certainty and predictability to technology companies like Facebook” by “enforcing choice-of-law provisions.” Dkt. No. 69 at 8. This makes little sense. If the chosen state’s interest in enforcing a choice-of-law provision alone were enough to trump the interest of the non-chosen state, Section 187 would largely be a nullity. And while California “certainly has a significant general interest in enforcing contracts executed.. .by its citizens,” Illinois has “a substantial, case-specific interest in protecting its resident^].. .from losing statutory protections.” Bridge Fund Capital Corp., 622 F.3d at 1004; see also Mazza v. Am. Honda Motor Co., 666 F.3d 581, 594 (9th Cir.2012) (“California’s interest in applying its law to residents of foreign states is attenuated.”).

Consequently, the Court declines to enforce the California choice-of-law provision and will apply Illinois law. Facebook’s summary judgment motion — specifically that the Court grant summary judgment for Facebook because the parties’ California choice-of-law provision precludes plaintiffs’ claims under Illinois law — is denied.

II. MOTION TO DISMISS

Facebook has also moved to dismiss on the ground that the statute excludes from the definitions of “biometric identifier” and “biometric information” (1) photographs and (2) any information derived from those photographs. Dkt, No. 69 at 10-11. Facebook says that its biometric data is derived exclusively from uploaded photographs and therefore it cannot be subject to BIPA. The motion is denied.

On a 12(b)(6) motion, the Court examines the sufficiency of the complaint under the plausibility standard set out in Bell Atl. Corp. v. Twombly, 550 U.S. 544, 570, 127 S.Ct. 1955, 167 L.Ed.2d 929 (2007). The Court accepts as true “[a]ll well-pleaded allegations of material fact in the complaint” and construes them “in the light most favorable to the non-moving party.” Faulkner v. ADT Sec. Servs., Inc., 706 F.3d 1017, 1019 (9th Cir.2013). The complaint survives a motion to dismiss when it alleges “enough facts to state a claim to relief that is plausible on its face.” Twombly, 550 U.S. at 570, 127 S.Ct. 1955. “A claim has facial plausibility when the plaintiff pleads factual content that, allows the court to draw the reasonable inference that the defendant is liable for the misconduct alleged.” Ashcroft v. Iqbal, 556 U.S. 662, 678, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009) (citing Twombly, 550 U.S. at 556, 127 S.Ct. 1955).

Under the plain language of BIPA, the complaint meets these pleading requirements. See BedRoc Ltd, LLC v. United States, 541 U.S. 176, 188, 124 S.Ct. 1587, 158 L.Ed.2d 338 (2004) (“[0]ur inquiry begins with the statutory text, and ends there as well if the text is unambiguous.”); State ex rel Pusateri v. Peoples Gas Light & Coke Co., 386 Ill.Dec. 674, 21 N.E.3d 437, 441 (Ill.2014) (“The primary objective of statutory interpretation is to give effect to the legislature’s intent, which is best indicated by the plain language of the statute itself.”). As discussed, BIPA regulates the collection, retention, and disclosure of personal biometric identifiers and biometric information by “[m]ajor national corporations,” among others. See 740 Ill. Comp. Stat. 14/5(b), (g); Section I.C.2(B), above. It defines “biometric identifier” as “a retina or iris scan, fingerprint, voice-print, or scan of hand or face geometry.” 740 Ill. Comp. Stat. 14/10. Plaintiffs allege that Facebook scans user-uploaded photographs to create a “unique digital representation of the face.. .based on geometric relationship of their facial features.” Dkt. No. 40 ¶ 23. That allegation falls within the scan of face geometry stated in the statute.

Facebook’s contention that the statute, categorically excludes from, its scope all information involving photographs, Dkt. No. 69 at 11-12, is unpersuasive. When interpreting a statute, the Court must “view the statute as a whole, construing words and phrases in light of other relevant statutory provisions and not in isolation.” People v. Gutman, 355 Ill. Dec. 207, 959 N.E.2d 621, 624 (Ill.2011); Util. Air Regulatory Grp. v. E.P.A., — U.S. -, 134 S.Ct. 2427, 2441, 189 L.Ed.2d 372 (2014) (“[T]he words of a statute must be read in their context and with a view to their place in the overall statutory scheme.”) (internal quotation marks and citation omitted). The statute puts photographs in the same list of excluded sources such as “writing samples, written signatures ... human biological samples used for valid scientific testing or screening, demographic data, tattoo descriptions, or physical descriptions such as height, weight, hair color, or eye color.” 740 Ill. Comp. Stat. 14/10. The statute’s focus is on newer technology like scans of face geometry, whose “full ramifications" are not known. Id. at 14/5(f). Read together, these provisions indicate that the Illinois legislature enacted BIPA to address emerging biometric technology, such, as Facebook’s face recognition software as alleged by plaintiffs, without including physical identifiers that are more qualitative and non-digital in nature. “Photographs” is better understood to mean paper prints of photographs, not digitized, images stored as a computer file and uploaded to the Internet. Consequently, the Court will not read the statute to categorically exclude from its scope all data collection processes that use images. And to read that categorical exclusion into the statute would substantially undercut it because the scanning of biometric identifiers is often based on an image or photograph. See Lebowitz v. City of New York, No. 12 civ. 8982, 2014 WL 772349, at *2 (S.D.N.Y. Feb. 25, 2014) (“Iris scans are high-resolution photographs of the pigmented portion of the eye”); see also Norberg v. Shutterfly, Inc. et al., 152 F.Supp.3d 1103, 1105, No. 1:15-cv-05351, 2015 WL 9914203, at *2 (N.D.Ill. Dec. 29, 2015) (denying defendants’ motion to dismiss premised on the same argument that BIPA excludes biometric identifiers derived from photographs).

Facebook argues that the only way to reconcile the statute’s inclusion of “scan” and exclusion of “photographs” is to read the word “scan” to mean in-person scan. Dkt. No. 69 at 12-13. But this cramped interpretation is not stated in BIPA and cannot be squared with the statute’s purpose. See Gutman, 355 Ill.Dec. 207, 959 N.E.2d at 624 (“The court may consider the reason for the law, the problems sought to be remedied, the purposes to be achieved, and the consequences of construing the statute one way or another.”); Gen. Dynamics Land Sys., Inc. v. Cline, 540 U.S. 581, 600, 124 S.Ct. 1236, 157 L.Ed.2d 1094 (2004) (interpreting statute in light of its “text, structure, purpose, and history”). The statute is an informed consent privacy law addressing the collection, retention and use of personal biometric identifiers and information at a time when biometric technology is just beginning to be broadly deployed. See 740 Ill. Comp. Stat. 14/5(f)-(g), 14/15. Trying to cabin this purpose within a specific in-person data collection technique has no support in the words and structure of the statute, and is antithetical to its broad purpose of protecting privacy in the face of emerging biometric technology-

At this pleading challenge stage, then, the complaint passes review. As the facts develop, it may be that “scan” and “photograph” with respect to Facebook’s practices take on technological dimensions that might affect the BIPA claims. Other fact issues may also inform the application of BIPA. But those are questions for another day. The Court accepts as true plaintiffs’ allegations that Facebook’s face recognition technology involves a scan of face geometry that was done without plaintiffs’ consent. Consequently, they have stated a plausible claim for relief under BIPA.

CONCLUSION

The motions to dismiss and for summary judgment are denied. A case management conference is set for June 15, 2016, at 1:30 p.m.

IT IS SO ORDERED.

Notes

1 . Both states look to the Restatement (Second) Conflict of Laws Section 187 for enforceability disputes, but do not apply it the same way. For example, Illinois uses a “pure morals and abstract justice” consideration, see, e.g., Potomac Leasing Co. v. Chuck's Pub, Inc., 156 Ill.App.3d 755, 759, 109 Ill.Dec. 90, 509 N.E.2d 751 (1987), but that does not appear to be a factor in California’s interpretation of Section 187.

Case Details

Case Name: In re Facebook Biometric Information Privacy Litigation

Court Name: District Court, N.D. California

Date Published: May 5, 2016

Citation: 185 F. Supp. 3d 1155

Docket Number: Case No. 15-cv-03747-JD

Court Abbreviation: N.D. Cal.