Court File and Parties
COURT FILE NO.: CV-20-648610-00CP DATE: 20210622
SUPERIOR COURT OF JUSTICE - ONTARIO
RE: Colin Loewenthal
AND:
Sirius XM Holdings, Inc., Sirius XM Radio, Inc., Sirius XM Canada Holdings, Inc., Sirius XM Canada Inc., Sirius XM Connected Vehicle Services Holdings Inc., Sirius XM Connected Vehicle Services Inc., SXM CVS Canada Inc. and Adswizz, Inc.
BEFORE: J.T. Akbarali J.
COUNSEL: Mark C. Canofari, Glyn Hotz, Darrel Hotz and Jodi Rhiger, for the plaintiff H. Michael Rosenberg and Stephanie Willsey, for the Canadian defendants Andrew Eckart, for the objector Michael McWha
HEARD: June 18, 2021
Proceeding under the Class Proceedings Act, 1992
ENDORSEMENT
Overview
[1] On this motion, I am asked to approve a settlement in this class proceeding. The proceeding itself was certified on consent against the defendants Sirius XM Canada Holdings Inc., Sirius XM Canada Inc., and Sirius XM CVS Canada Inc. (the “Canadian defendants”) for settlement purposes only on March 26, 2021: 2021 ONSC 2318. For the reasons below, I approve the settlement.
Brief Background
[2] The plaintiff commenced this action by way of Notice of Action on October 1, 2020. The claim alleges that the defendants unlawfully collect, maintain, store and use data and personal information belonging to the plaintiff and class members who purchase or service new or pre-owned vehicles which come equipped with a SiriusXM satellite and internet radio services. The action seeks damages of $500,000,000, and alleges a number of causes of action, including intrusion upon seclusion, negligence, breach of contract, and statutory causes of action, to name but a few.
[3] Shortly after the claim was served, the Canadian defendants indicated that they intended to move for summary judgment. The non-resident defendants were considering a jurisdictional motion. There was disagreement between the parties about whether the summary judgment motion should proceed before or after a certification motion. As a result, on December 1, 2020, I released a timetabling endorsement scheduling a sequencing motion. I also addressed the intended amendment of the plaintiff’s claim which, in part, was expected to clarify which defendants were properly part of the proceeding.
[4] I was subsequently advised that the parties had reached an early settlement agreement. Under the first iteration of the settlement agreement, the defendants agreed to make a cy-près payment of $50,000 to the Automobile Protection Association (“APA”) – an Ontario-based consumer watchdog – to fund a report on best practices to address privacy concerns when purchasing a vehicle. In exchange, the defendants were to receive a release from class members. Class counsel was to receive $100,000 in fees, and the representative plaintiff and another class member who had been involved in instructing class counsel were to receive a $5,000 honoraria.
[5] I certified the proceeding on consent for purposes of settlement, but I expressed some concerns about the value of the settlement to the class, especially because it was half of the proposed class counsel fees. By the time of the settlement approval hearing, the parties had amended the settlement agreement to provide for a payment of $150,000 to the APA, and $75,000 to class counsel. Class counsel fees and the honoraria are not before me on this motion, but will be the subject of a further motion.
[6] The class that has been certified for the purposes of settlement in this case is:
Any person in Canada who purchased, leased, or obtained servicing for a new or preowned vehicle equipped with a satellite radio receiver or similar device whose Personal Information was collected, accessed, maintained, stored, lost, used, or disclosed directly or indirectly by one or more of the Defendants.
[7] The parties undertook a notice campaign, including 5.3 million emails providing direct notice. The notice campaign resulted in so many enquiries to class counsel that I approved a variation to the notice program, so that the form of notice would better answer common questions that class members had raised. The parties retained Ricoh eDiscovery to review the thousands of responses received as a result of the notice program. Their report is before me and indicates that 144 class members appeared to consent to the settlement, while 82 objected. Another 2,045 class members clearly opted out of the settlement, while an additional 4,003 may wish to opt out. The parties have indicated that they have treated all objectors as opt-outs, and have treated the 4,003 who may wish to opt out as opt-outs.
[8] At the settlement approval hearing one objector made submissions through counsel from the Class Action Clinic at the University of Windsor. I note that the settlement approval hearing was well-attended. It proceeded by zoom, and including counsel and court staff, there were over 80 attendees.
Legal Principles Applicable to Motions to Approve a Settlement in a Class Proceeding
[9] Under s. 27.1(1) of the Class Proceedings Act, 1992, S.O. 1992, c. 6 (“CPA”), a proceeding brought under the CPA may only be settled with court approval. The court shall not approve a settlement unless it determines that the settlement is fair, reasonable, and in the best interests of the class: s. 27.1(5) CPA; Nunes v. Air Transat A.T. Inc., 2005 CarswellOnt 2503 (S.C.J.), at para. 7. The burden lies on the party seeking approval: Nunes, at para. 7.
[10] Public policy favours the resolution of complex litigation: Nunes, at para. 7; Amoco Canada Petroleum Co. v. Propak Systems Ltd., 2001 ABCA 110 at para. 27, (2000), 200 D.L.R. (4th) 667 (Alta. C.A.).
[11] Settlements need not be perfect; they are compromises: Lozanski v. The Home Depot, Inc., 2016 ONSC 5447, at para. 71. To find that a settlement is not fair and reasonable, it must fall outside a range of reasonable outcomes: Nunes, at para. 7; Haney Iron Works v. Manufacturers Life Insurance, (1998), 1998 CanLII 3085 (BC SC), 169 D.L.R. (4th) 565 (Ont. S.C.), at para. 44. An objective and rational assessment of the pros and cons of a settlement is required: Mancinelli v. Royal Bank of Canada, 2017 ONSC 2324, at para. 38. There is a strong presumption of fairness when a proposed class settlement, which was negotiated at arms-length by counsel for the class, is presented for court approval: Nunes, at para. 7.
[12] A court must be assured that the settlement secures appropriate consideration for the class in return for the surrender of its litigation rights against the defendants: Nunes, at para. 7. However, it is not the court’s function to substitute its judgment for that of the parties or attempt to renegotiate a proposed settlement. Nor is it the court’s function to litigate the merits of the action, or, on the other hand, to rubber-stamp a settlement: Nunes, at para. 7.
[13] When considering whether to approve a negotiated settlement, the court may consider, among other things: (a) the likelihood of recovery or likelihood of success; (b) the amount and nature of discovery, evidence or investigation; (c) the proposed settlement terms and conditions; (d) the recommendation and experience of counsel; (e) the future expense and likely duration of litigation and risk; (f) the recommendation of neutral parties, if any; (g) the number of objectors and nature of objections, if any; (h) the presence of good faith, arm’s length bargaining and the absence of collusion; (i) the degree and nature of communications by counsel and the representative parties with class members during the litigation; and (j) information conveying to the court the dynamics of and the positions taken by the parties during the negotiation: Lozanski, at para. 73; Nunes, at para. 7.
Is the proposed settlement fair, reasonable, and in the best interests of the class?
[14] There is no doubt that the settlement proposed here was reached in good faith, through arms-length negotiation, and without collusion. It was apparent at the hearing that there are matters on which class counsel and defendants’ counsel continue to disagree.
[15] There is also no doubt that counsel for the class and for the defendants are experienced counsel. Moreover, class counsel consulted with experts including a lawyer from the Ontario Motor Vehicle Industry Council, a computer science professor, two professors in the field of Privacy by Design, and Data Integration, automative dealership owners, and counsel in American proceedings brought against the defendants. In these circumstances, there is a strong presumption that the settlement is fair: Nunes, at para. 7.
[16] But the proposed settlement also differs starkly from the relief sought in the claim. The claim seeks $500,000,000 in damages, and the settlement, reached very early in the litigation, before discoveries, provides for a cy-près payment of just $150,000.
[17] The discrepancy between what is sought in the claim and what is proposed as a settlement can be explained by the significant risks raised by this class proceeding.
[18] For example, the evidence before me on this motion indicates:
a. Sirius XM Canada Inc. (“SiriusXM”) has agreements with original equipment manufacturers (“OEMs”) and vehicle dealers to provide its satellite radio services to users. OEMs install a satellite radio receiver in their vehicles at their factories. When a person buys, leases, or sometimes even just services a vehicle at a dealer, they may receive a free trial of SiriusXM satellite radio, and may thereafter elect to convert the trial to a paid subscription.
b. SiriusXM has agreements with OEMs who in turn have agreements with dealers. SiriusXM’s agreements with OEMs require that their dealers take reasonable measures to obtain appropriate consent to collect their clients’ personal information and share it with SiriusXM, which uses it to offer users a paid subscription.
c. An example of a customer consent form in the record indicates that customers are asked to sign a form with a clear heading, “Consent to Sharing of Personal Information,” under which there are three boxes a customer can check, including “I consent to my personal information being disclosed to satellite radio providers.”
d. The record also includes a sample brochure from a dealer (in this case, Subaru, from which the representative plaintiff purchased his vehicle) and SiriusXM, that markets the satellite radio service to customers. It indicates that once the three-month trial subscription ends, the customer has a choice of different packages. The brochure includes an indication that the customer will be contacted about the service: “Watch out for our Welcome Communication with your username and password.”
e. There is also a page with “Important Information about Your Subscription,” that includes a statement, in bold, that “We may get your contact information from your dealer, automaker or other third party, and contact you by mail, phone or email to discuss subscription options.” It then provides a link to SiriusXM’s privacy policy.
f. There is another example of a fold-out brochure in the record, which includes the same statement about how SiriusXM gets customers’ contact information.
g. The record indicates that SiriusXM maintained the following data on the representative plaintiff: his name, his address, his phone number, his email, his plan description (the free three-month trial of SiriusXM satellite radio), the start and end date of his plan, and the date he bought his vehicle. Mr. Loewenthal gave evidence that he does not consider any of this information to be private.
h. The evidence also indicates that SiriusXM’s receipt and use of Mr. Loewenthal’s information was typical of its practices, and that it does not obtain sensitive personal information from trial subscribers.
i. SiriusXM’s 2017 privacy policy (the one in effect when the plaintiff bought his vehicle) indicates:
…When you purchase a SiriusXM satellite radio (either as a stand-alone device or integrated into a vehicle), our partners will often ask for your consent to collect your information and to share it with certain third-parties, including SiriusXM. Where our partners have sought such consent, we may receive your name, address, telephone number, your satellite radio’s Radio ID or ESN, and the vehicle’s VIN, make, model, and whether you purchased or leased the vehicle. In certain instances we may also receive your email address and demographic information.
j. The record indicates that SiriusXM maintains the class members’ data on secured servers. There is no known data breach, or data hack of customers’ information. There is no evidence that SiriusXM ever lost custody of customers’ personal information. There is no evidence that customers’ information was improperly handled or stored.
k. There is no evidence that customers’ information was sold to, or provided to, third parties.
l. Class counsel is not aware of any class member who can provide demonstrable evidence that they suffered actual harm as a result of the defendants’ collection or use of their personal information.
m. Class counsel investigated SiriusXM’s response to Mr. Loewenthal’s request for his personal information, reviewed the defendants’ policies and procedures, developed a good understanding of SiriusXM’s business and the measures it had implemented to comply with applicable privacy legislation. Having done so, they found “no convincing evidence that SiriusXM’s collection of users’ personal information broke any laws, that SiriusXM lost or was careless in handling users’ personal information, or that SiriusXM sold users’ personal information to third parties.”
n. Class counsel recommends the settlement.
o. The representative plaintiff, who is an IT expert, has been involved in the proceeding in a significant manner, and also recommends the settlement.
[19] The Canadian defendants argue that the evidentiary record demonstrates that they have strong defences on the facts, including both having obtained express consent to collect class members’ information, and implied consent to collect class members’ information. They note that the Personal Information and Electronic Documents Protection Act, S.C. 200, c. 5, Sched. 1 (“PIPEDA”) has as its purpose balancing privacy interests with reasonable commercial interests, and permits organizations to obtain implied consent for collection of less sensitive personal information. Moreover, implied consent is available even for presumptively sensitive categories of personal information: Royal Bank of Canada v. Trang, 2016 SCC 50. In any event, there is no evidence of any actual damages suffered by any class member.
[20] The defendants liken this case to Emond v. Google LLC, 2021 ONSC 302, which dealt with a claim of privacy protection and non-consensual transmission or collection of phone data by Google. In that case, Morgan J. approved a $1 million cy-près settlement, noting the weakness of the case for certification, and finding that Google had offered to pay” $1 million more than it likely would have had to pay had the matter been litigated to trial.”
[21] Similarly, in Cass v. WesternOne Inc., 2018 ONSC 4794, Glustein J. approved a cy-près settlement where the risks of lack of success were overwhelming.
[22] The plaintiff’s submissions on the risks in the litigation were a little confusing. On one hand, counsel submitted that the plaintiff largely accepted the defendants’ submissions. However, Mr. Glyn Hotz took some time in submissions to review where he believed there was strength in the plaintiff’s case. To some extent, those submissions relied on assumptions of facts that were not before me in evidence. To Mr. Hotz, the biggest risk was that the individual issues might overwhelm the common issues and lead to difficulty in certifying the case. Mr. Hotz argued that given the multitude of relationships with individual dealers – each dealership being individually owned – each individual class member might have difficulty proving that any improper collection of personal information was the responsibility of the Canadian defendants as opposed to the individual dealers.
[23] Class counsel also noted some changes in the law that occurred after the Notice of Action was filed. In Setoguchi v. Uber B.V., 2021 ABQB 18, the court found that nominal damages are unavailable for negligence or breach of contract. Rather, privacy claims will only be viable where claimants can show that the breach of their privacy caused actual harm. Mr. Darrel Hotz, class counsel, deposed in his affidavit that “the evolving jurisprudence has made it difficult to advance a class proceeding in these circumstances.”
[24] I agree that the individual issues pose a significant risk to certification. I also agree that the available evidence establishes that there is significant risk to the plaintiff on a summary judgment motion. I note class counsel’s factum, which states:
To possibly achieve success at the common issues trial as against SiriusXM, Class counsel determined that the Plaintiff and Class Members will have had to have overcome significant hurdles, including showing that the Data and Personal Information was not in the public domain, or was so vast that it amounted to a privacy breach (novel in law in Canada), that it was collected without consent or free and informed consent, and/or that it was maintained, stored and used improperly, and/or that SiriusXM’s Privacy Policy was inadequate. There was a risk of failing either at a summary judgment stage or at certification. In fact there was only marginal support for any of these propositions. In short, this case was hanging by a thread, but the thread persisted.
[25] The action is at an early stage; pursuing it would be expensive and time-consuming. If the settlement is not approved, the proceeding will return to square one. The sequencing motion, summary judgment motion, and certification motion all loom large. The costs of the proceeding, and the costs exposure of proceeding, are significant, especially in light of the serious risks in the plaintiff’s case.
[26] Although there have been no discoveries as of yet, class counsel’s affidavit explains the investigations and due diligence counsel has undertaken to assess the prospects of success. I am satisfied that they have conducted a reasonable assessment of risks despite the fact that no discoveries have taken place.
[27] The settlement does not provide any direct benefits to the class. Counsel argues that the settlement is tailored, and provides indirect benefits to the class. Cy-près settlements have been ordered where (i) it is not practical to distribute the benefits in any other manner, (ii) a direct distribution to the settlement class would be uneconomic considering the modest damages and the fact that there is no cost effective way of locating the settlement class members, determining if they suffered damage and, if so, establishing their loss, and (iii) the distribution is directly related to the issues in the lawsuit and will directly benefit people in similar circumstances to the class members: WesternOne, at para. 91.
[28] The objector takes some issue with the evidence about the proposed distribution to the APA, arguing that someone from the APA should have filed an affidavit directly. While that might have been preferable, the affidavit of Darrel Hotz describes the proposed cy-près distribution in detail, and explains that class counsel had significant involvement in its development.
[29] The proposal is that the APA will work with Professor Debbie Roberts and Mr. Iny to develop a teaching module for students and automotive industry stakeholders. Professor Roberts’ PhD dissertation was entitled “A Kantian Defense of Regulatory Law and Penalty: Consumer Protection in the Canadian Automotive Industry.” She has worked with the Ontario Motor Vehicle Industry Council, which regulates dealerships in Ontario, and been active as a member of its Discipline and Appeals Committee. Mr. Iny has four decades of experience as an expert in the auto industry and in consumer protection, and has been a speaker at the Automotive Business School of Canada.
[30] The module will be available on the APA’s website. Class counsel will assist with the interpretation of privacy laws on a pro bono basis.
[31] I accept that this proposed cy-près distribution is tailored to the nature of the allegations in the statement of claim. I also accept that it is not practical, given the size of the settlement, to distribute it to the class. It is not known exactly how large the class is, but it is at least over 2 million people, as SiriusXM has over 2 million subscribers. As I have noted, over 5 million direct emails were sent to potential class members. I infer that direct distribution of proceeds to the class members would be uneconomic and impractical. Moreover, in the absence of any evidence of actual harm to the class members, the indirect benefit of the proposed cy-près distribution to the class members and others in similar circumstances furthers the goal of behaviour modification in the automotive industry by educating stakeholders about their privacy rights and obligations.
[32] The objector argues that the release that the settlement agreement provides is too broad. He states that the class is being asked to give up something of value, and the indirect benefits from the proposed cy-près distribution provide insufficient value in exchange.
[33] I agree with the objector that, when assessing a settlement, the court must direct its mind to the scope of the release that is proposed: Berg v. Canadian Hockey League, 2020 ONSC 6389, at paras. 52-53. Counsel for the Canadian defendants was frank that, without a release, the defendants would have wanted to proceed with their summary judgment motion, so as to avoid a proliferation of privacy-based claims to which it would have to respond. The release is an integral part of the deal reached between the parties.
[34] However, I disagree with the objector that the release is drafted too broadly in this case. It is a broad release, but it responds to the extraordinarily broad claims made in the statement of claim. Unlike Berg, where the proposed release raised the concern that it could bar claims in other existing class actions, the release proposed in this case does not extend beyond the claims asserted in the statement of claim.
[35] The objector argues that the release extends to all claims “known or unknown, suspected or unsuspected, foreseen or unforeseen…” and that this is too broad. I agree that if that was the breadth of the release it would be too broad. But those words are limited by words that follow in the definition of “Released Claims” contained in the Settlement Agreement. For example, they are limited by the phrase “relating to any conduct alleged in the subject matter of this Action, or which could have been alleged pertaining to the subject matter in the Action.” Those words circumscribe the subject matter of the release to the claims related to the allegations around the defendants’ collection, storage, and use of the class members’ personal information.
[36] The release is also limited temporally, “from the beginning of time to the date hereof.” While “from the beginning of time” seems astonishingly broad at first blush, one must recall that any claims existing at the beginning of time would have long ago expired by operation of the limitations period, or alternatively, laches. The defendants themselves did not exist at the beginning of time. What is important in the temporal limit is that it does not go into the future. The release only releases claims “to the date hereof.” The settlement agreement is dated as of February 25, 2021.
[37] The objector argued that the settlement agreement operated as a form of indemnity for the defendants, and would operate as a disincentive for it to modify its behaviour going forward. I have difficulty with this argument. First, there is absolutely no evidence of a data breach in the past, or of any actual harm suffered by any class member. Second, there is no evidence of wrongdoing on the part of the defendants. Without evidence of wrongdoing, what behaviour requires modification? Finally, even assuming the defendants could improve their practices around the collection, storage and use of personal information, why would a time-limited release, that does not cover future behaviour, discourage them from modifying their behaviour? The suggestion that it would is not only speculative, it does not make sense. The defendants are exposed to liability should they fail to meet their privacy-related obligations to their customers in the future. If anything, their awareness of such exposure has been heightened by the experience of dealing with this class proceeding. Knowing that their current and future practices are open to scrutiny provides an incentive for them to modify their current and future behaviour if it needs modifying.
[38] The objector also argued that the release should only extend to the certified claims, and not all the claims that were alleged or could have been alleged in the proceeding. I disagree. In a class proceeding, all the claims are in issue, not only those that are certified as common issues. When an issue is certified as a common issue, it is resolved in a manner common to the class. Other issues raised in the proceeding remain to be resolved individually. Moreover, it is common for releases to extend to issues which could have been alleged in the claim, in order to capture the substance of the issues that are being settled.
[39] Settlements are a compromise. It is unrealistic to expect a settling defendant to be content to walk away without a meaningful release.
[40] The objector argues that the risks in the proceeding are an indication that the plaintiff should discontinue its claim, and ask the court to allow him to do so on a no costs basis. In my view, this argument is unrealistic. The representative plaintiff and class counsel have invested significant time in the pursuit of this action, albeit it remains at an early stage. I accept that the claim was brought in good faith. In addition to the proposed cy-près distribution, the claim has had the beneficial effect of raising awareness among class members, and perhaps others who read the notices published in the newspaper, about data collection and privacy-related issues. There is now a deal on the table which protects the representative plaintiff from adverse costs awards, and provides an indirect benefit to the class and others in similar circumstances. In exchange, the class is being asked to release claims with a very low chance of success. There is no other class member who is seeking to step into the representative plaintiff’s shoes, with the attendant work and risk, and advance the claims.
[41] Moreover, it is only those class members who did not object who are releasing their claims. Because anyone who expressed concern or objected is taken to have opted out, the objector is not releasing his claims. Because the certification of the class was for the purposes of settlement only, the class members here are able to assess the settlement and decide for themselves if they want to be part of the class.
[42] The objector also argued that the opt-out form was confusing. I do not agree. I approved the notice program twice. The fact that over 6000 people objected indicates that the opt-out form was not confusing. Moreover, some people opted out, not by using the prescribed form, but by sending an email, and those opt-outs were also registered.
[43] The direct notice campaign was successful; as I have noted, thousands of communications were received, and over 6,000 people are registered as having opted out. At the same time, the number of opt-outs is small compared to the direct notice campaign. Leaving aside those who may have learned about the settlement from the notices published in newspapers, the number of opt outs represents 0.0011% of people who received direct notice of the settlement.
[44] The evidence indicates that there were common reasons cited by those who were opting out. Some opted out because they did not believe they were customers of SiriusXM. Others wanted to maintain their rights to pursue the defendants should there be reason at a later stage. Still others believed there should be direct compensation to class members, an objection I have already addressed through my examination of the risks of this litigation. Others thought the defendants should face stiffer consequences, an objection that is also addressed by my examination of the risks of this litigation. Others objected to class counsel’s fees, an objection that will be addressed in the future at another motion. Finally, other miscellaneous reasons included those who did not understand the notice, did not want to give up their time, or risk costs, or pay fees.
[45] Overall, the objectors are a small proportion of the class members, and many of the reasons for the objections have been addressed in my analysis of the risks of the claim, and benefits of the settlement.
[46] Like Perell J. in Lozanski, at para. 74, in this case, I would have approved a discontinuance of the proposed class action with or without costs, and without any benefits achieved by class members due to the risks in the litigation. In these circumstances, I conclude that the proposed cy-près distribution is a fair and reasonable settlement that is in the best interests of the class.
Conclusion
[47] The settlement is approved. It is fair and reasonable, and in the best interests of the class.
[48] Counsel has filed an order which must be finalized. In particular, it is necessary to determine how best to identify those people who have opted out of the class in a way that protects their privacy interests. I shall convene a case conference for that purpose, and to address next steps.
J.T. Akbarali J.
Date: June 22, 2021