Thank you, Mr. Chair.
I am here alone. Janet Lo, my co-counsel, sends her regrets. She's in a lock-up for CRTC on Bell-Astral.
The Public Interest Advocacy Centre is a non-profit organization that provides legal and research services on behalf of consumer interests, and in particular vulnerable consumer interests, concerning the provision of important public services. We have been deeply involved with the Personal Information Protection and Electronic Documents Act, PIPEDA, from a consumer perspective since its passage. We have published several recent reports: one on children's privacy online, one on a do-not-track list, and one on data breaches.
I've given the clerk a copy of references to those and summaries.
We're here today to talk about the immediate future of privacy. It is largely to be defined by services such as social networks. But social networks provide challenges to our concept of personal information and the commercial interests that are involved with that.
PIAC recently brought a complaint to the Office of the Privacy Commissioner of Canada under PIPEDA against Nexopia.com Inc., a social network based in Alberta and largely aimed at a teen audience. This real-life example illustrates the challenges of dealing with privacy and social networks, and unfortunately the inadequacies of PIPEDA to deal with improper privacy practices, even those where the improprieties involve children and teens.
PIAC alleged that Nexopia provided no comprehensible descriptions of the collection, use, and disclosure of the personal information of their largely underage users. We said that the company did not adequately detail its disclosure of information to advertisers, nor did it adequately detail how it used this information to serve up targeted teen ads. We complained that the default settings for personal information like gender, age, location, and pictures were open to the Internet—that is, not even closed to members of the site—and that this was unreasonable and even dangerous for the young users of the site. Finally, we noted that Nexopia appeared to keep personal information forever, even if an account were deleted.
The Privacy Commissioner upheld all our complaints. That was February 2012, some two years after we filed it.
Regarding default settings, the Privacy Commissioner wrote, in part:
We do not consider making portions of a user's profile available to anyone on the Internet to be consistent with users' reasonable expectations, particularly when a user has clearly indicated his or her preference to share information on a more limited basis.
However, Nexopia has said to the Privacy Commissioner that they will not implement the four recommendations related to retention of data. The Privacy Commissioner has had to go to Federal Court to enforce her findings. Why?
First, the Privacy Commissioner has no order-making power. She has no fining power. Social networks that judge privacy findings too inconvenient or expensive, it appears, can continue to operate in a privacy-violating manner.
Second, the refusal reveals the real nature of social networks: they are financed by personal information. Asking a social network to destroy data appears to them like removing an asset from the balance sheet.
The Privacy Commissioner's trip to Federal Court will show if business purposes or the personal privacy of individuals is paramount under PIPEDA. However, the larger issue for you at this committee is how to help design laws to avoid this type of conflict from arising in the first place, particularly in the fast-moving social networking and online space.
Now I'll move to Bill C-12 and breach notification.
LinkedIn and eHarmony suffered large data breaches this spring. Social networks are now major targets of hackers, and there is a risk of exposure of personal information that is not intended for general viewing from these websites. This is in addition to the leaking of personal information from websites noted by the Privacy Commissioner at the end of September in a recent study.
Bill C-12 is intended to amend PIPEDA to provide for data breach notification. However, it does not succeed. It allows the company suffering the breach to make the determination of whether the breach is material enough to even report to the Privacy Commissioner. Part of that determination is an assessment, again made by the company of itself, of whether the cause of the breach or a pattern of breaches indicates a systemic problem.
It's extremely unlikely, in our view, that any company, but particularly a social network that trades in data, will declare that it has a systemic problem with data breaches and data handling that leads to breaches.
Bill C-12 is asking companies to declare that they, in effect, are negligent. As a result, we confidently predict that under Bill C-12 a social network or other online company will almost never notify the Privacy Commissioner of a breach that has not otherwise been made public. Companies are expected to determine whether to report data breaches directly to the consumers as well. They must determine if it is reasonable in the circumstances to believe that the breach creates a real risk of significant harm to the individual.
First, this threshold is very high. It's higher than U.S. state law requirements and it's unrealistic. It's difficult to predict how personal information will be misused.
Secondly, Bill C-12 ignores the blindingly obvious incentive for companies to find no such risk to individuals and avoid notification and its cost. As a result, we confidently predict that under Bill C-12, social networks and other online media companies will almost never notify individuals of a breach that has not otherwise been made public.
There is another model in Canada for data breach laws: the Alberta Personal Information Protection Act. In Alberta, all breaches must be reported to the Privacy Commissioner of Alberta, on pain of fines. The Alberta Privacy Commissioner then determines if the breach is serious enough to notify individuals on a test of potential for any harm.
PIAC studied public attitudes to data breach notification in focus groups in 2011. Overwhelmingly, participants preferred the Alberta-type model to leaving companies to make this decision. We urge this committee to express these concerns about breach notification under Bill C-12 in its report.
I will turn now to privacy policies. Social network privacy policies are “take it or leave it” contracts. The burden of determining what is done with personal information is borne by the user. Yet social networks regularly rely on the consent of users to justify practices and point to the use of the site as the equivalent of consent to the entire privacy policy.
It's PIAC's view that this legal fiction is in fact used in place of informed consent in many social networks. Users simply do not read all the policy, and if they do, they do not understand it. Why is this? This is because major social networks define “personal information” in confusing ways, and none of them define it in the way it is defined in PIPEDA.
Many define personal information as personally identifiable information, which, as you recognize, is a U.S. legal concept. Recently, many larger websites have dropped any definition at all of personal information, only to give examples of treatment of certain data elements like gender or age. The clerk also has a copy, which should have been distributed to you, of wording of privacy policies that we're talking about.
This non-definition of personal information matters because users reading the privacy policy are not able to understand their real rights under PIPEDA in order to launch a complaint or to bring the company into compliance or even to contact the company.
The Privacy Commissioner appeared before this committee and stated that social networking sites do not do a sufficient job of explaining their use of personal information. She said she doubts in these situations that the social networking site has real consent. We think the Privacy Commissioner is right. But the complaint mechanism under PIPEDA is very poor enforcement. She needs order-making and fining power.
PIAC suggests, however, that given the challenges of big data collection by social networking and other online businesses, this committee go further and consider a full enforcement framework such as that for the do-not-call list for companies flouting Canadian privacy law.
I'm going to close with some forward-thinking ideas on social networking and privacy.
First of all, there are many related entities dealing with personal information created at social networking sites in order to monetize that information through advertising and other methods. This committee should study these relationships and consider rules for revealing related parties in personal information trafficking akin to those rules in securities law to bring increased transparency to data flows in social networking sites and marketing companies.
Secondly, the committee should consider a national do-not-track list.
Thirdly, the committee should study the nexus between privacy and competition law, and whether the Competition Bureau actually has a role to play in addressing privacy concerns and where a merger or other practice can reduce competition. For many online markets, competition for eyeballs depends on the currency of personal information or the value of big data.
PIAC thanks the committee for this opportunity to speak. We are happy to answer questions
in both English and French.
Thank you.