On March 9, 2016 the Department of Innovation, Science and Economic Development Canada released a discussion paper on the new data breach regulations being proposed. The Ministry is accepting public submissions until May 31, 2016 on the proposed Data Breach Notification and Reporting Regulations.
The Digital Privacy Act (also known as Bill S-4), which received Royal Assent on June 18,2015, amended Canada’s private sector privacy law, the Personal Information Protection and Electronic Documents Act (PIPEDA).
The amendments made several important changes to PIPEDA, including adding the requirement that private-sector organizations notify Canadians when their personal information has been lost or stolen, and they have been put at risk of harm as a result. Also added was a requirement to report these potentially harmful data breaches to the Office of the Privacy Commissioner of Canada (“OPC“) (for a more detailed analysis of the changes, see our previous blog post here).
However, these new data breach requirements in PIPEDA have not yet come into force, pending the formulation and passage of regulations.
The regulations are now being considered. The Ministry discussion paper is intended to solicit stakeholder input and views on these regulations. Following this consultation process, the government will publish draft regulations in Part I of the Canada Gazette for public comment and consultation. Final regulations will be published in Part 2 of the Canada Gazette, and PIPEDA’s new data breach provisions will then be brought into force. No timeline has been publicly set for this process.
The discussion paper not only solicits comments, it identifies issues that may arise in respect of certain regulatory approaches. In that regard, it is instructive to businesses which, even if they aren’t considering providing comments, will want to be aware of the directions in which the government may be considering heading.
Notably, some of the questions being explored are: whether certain industries should be subject to specific regulations (see our previous blog post with respect to the auto sector); whether and how information information-sharing among organizations should be required; and what record keeping requirements should be imposed on organizations.
Elements of the Regulations
- Determining Real Risk of Significant Harm
Under the new PIPEDA data breach amendments, organizations that become aware of breach of their security safeguards must determine whether or not the breach poses a “real risk of significant harm” to an individual whose personal information was involved in the breach. This determination acts as a trigger for a cascading series of notification and other decisions.
The nature of the analysis required is vague. Section 10.1(8) requires that, at a minimum, an organization consider the sensitivity of the personal information involved in the breach and the probability that the information, if misused, will result in harm. The government intends to specify in teh regulations additional factors that are relevant in making this determination.
Other factors that may be considered are context, the nature of the information at issue, the probability the data could be misused, whether the data was encrypted and so on. The discussion paper explores the approaches taken to this issue by several jurisdictions (Alberta’s breach reporting assessment tool, by the OPC’s own existing voluntary guidelines, the proposed U.S. Personal Data Notification and Protection Act, and the European Union’s ePrivacy Directive).
The Questions for Discussion identified by the government are:
Question 1: Is it necessary to identify additional risk-assessment factors in the regulations? Or are the factors listed in the legislation sufficiently clear?
Question 2: If additional factors should be prescribed, what are they?
Question 3: Should the regulations specify that the risk to individuals can be presumed to be low in circumstances where appropriate encryption has been used? How should that appropriate level of protection should be defined?
Adding additional risk factors may increase costs to organization as they do the required analysis and may risk bogging down the breach management process; it may also result in little additional benefit to persons affected by a breach. The point about encryption is an interesting one – potentially a kind of “get out of jail free” card for organizations which encrypt personal information. In practice, however, few organizations routinely do this as it can be cumbersome and lead to data processing issues.
- Report to Commissioner – Form and Content
If an organization concludes that a breach of security safeguards poses a real risk of significant harm, section 10.1(1) of PIPEDA requires that the organization report the breach to the OPC. The government has the authority to list the types of information that must be included in such a report and to specify a particular form and manner for such reports.
Typically such a report will include the date and location of the breach and date of its discovery, a description of the incident; the cause of the breach; an estimate of the number of individuals affected and the relation of those individuals to the organization (e.g., employee, customer); the type of information involved; measures taken by the organization to contain the breach; and whether anyone else has been notified of the incident (e.g., affected individuals, law enforcement) and when (see for instance, the Alberta Personal Information Protection Regulation or the existing OPC voluntary Privacy Breach Incident Form).
This question of form and content is not simply one of process. In a large breach of sensitive personal information, regulatory and litigation risk means that organizations will want to tightly control the information they are releasing. The requirement to say one thing to one regulator and a different thing to another means a lot of time will be spent reviewing and revising multiple reports to ensure consistency. In the US, where multiple inconsistent state-by-state reporting regimes exist, this is a cumbersome, time consuming process that increases risk for businesses.
The government appears to be aware of this, and has asked the following questions:
Question 4: (a) Is there any information required under the existing OPC voluntary report form that should not be included in a mandatory report? If so, what?; (b) Is there additional information that should be required? If so, what?
Question 5: Should reports to the Commissioner contain an assessment by organizations of the type of harm that may result from a breach and the likelihood of that harm occurring?
Question 6: To what extent should the completion and validation of all elements of the report to OPC be required for an organization to be considered in compliance with the reporting requirement?
Question 7: Should the regulations require organizations to update the OPC in circumstances where the information provided in the original report is discovered to be inaccurate, incomplete or has changed?
Question 8: Should organizations be required to report to the OPC in written format only (electronic or hardcopy) for greater efficiency? If not, why?
Question 9: Should a secure, electronic means of reporting data breaches to the Privacy Commissioner be established?
- Notification to Individuals – Content
Section 10.1(3) of PIPEDA requires organizations to notify individuals of any breach involving their personal information that poses a real risk of significant harm. Section 10.1(4) provides that the notification must contain “sufficient information” to ensure the individual understands the risks posed by the breach, and what steps, if any, he or she can personally take to reduce or mitigate the harm.
This raises similar issue to those in #2 above, with the added risk that (at least for certain types of breaches) there is a high risk that a notification to an affected individual will be forwarded on to others, including the media and regulators. As a result, organizations will likely want some flexibility in what should be included in a mandatory notification to individuals.
The government has asked the following questions in this regard:
Question 10: Is it necessary for the regulations to identify specific information to be included in notifications to individuals or is the legislation sufficiently clear?
Question 11: If specific information should be prescribed, what information should it be?
- Notification to Individuals – Form and Manner
The government places a great degree of emphasis on the notification of individuals, calling it the “single most important aspect of the framework”. Ensuring an individual receives a data breach notification that is understandable ostensibly allows an individual to identify risks and take steps to protect themselves from such risks if necessary.
This is reflected in section 10.1(5) of PIPEDA which requires that the notice be communicated directly to an affected individual in a manner that ensures it is not confused with “junk mail” or hidden in other communications material.
In some circumstances however, direct notification to affected individuals may not be feasible. This is common where an organization collects personal information other than address, phone number or email. In such circumstances, indirect notification (e.g. via newspaper ads or web site notices) may be necessary.
The manner in which an organizations communicates with its customers or other affected individuals is a critical one – businesses are best positioned to understand the channels that they and their customers use. While few businesses would likely want to risk the reputational impact that can accompany newspaper ads or other “broadcast” types of notification, the costs of direct notification should not be underestimated. If the only contact information an organization has is address information, letters may be the only mechanism for direct notification – at enormous cost.
Email notification carries its own risks. While generally less expensive, faster and more convenient than other forms, such electronic notifications could create secondary issues such as non-compliance with Canada’s Anti-Spam Legislation, the risk that such communications can be forwarded directly to media, and the ease at which email notifications can be co-opted by fraudsters to target customers and affected individuals for sophisticated phishing attacks.
The government has asked for input on the following questions:
Question 12: What methods of communication should be permitted for direct notification to individuals?
Question 13: Should the regulations set any conditions and/or limitations on the use of any method of direct communication? If yes, what conditions and/or limitations?
Question 14: Should the regulations set-out specific requirements for notifications to be conspicuous and distinct from other communications?
Question 15: In what circumstances should organizations be permitted to indirectly notify individuals of a data breach?
Question 16: If cost is a consideration in determining whether indirect notification is permitted, how should the regulations establish the appropriate threshold?
Question 17: For indirect notifications to individuals, what methods of communication should be permitted?
Question 18: Should the regulations place any conditions and/or limitations on the use of any method of indirect communication? If yes, what conditions and/or limitations?
- Notification to Other Organizations
The government has stated that the primary objective of the new data breach reporting and notification framework in PIPEDA is to prevent or mitigate the potential harm to individuals resulting from a breach. It acknowledges that, in its view, some breaches may require organizations to notify other organizations that are in a position to reduce or mitigate the risk of harm. As a result, Section 10.2(1) requires organizations to notify third parties of a potentially harmful data breach if the organization making the notification believes that the third party may reduce or mitigate the potential harm.
In addition, Section 10.2(1) gives the government the authority to prescribe specific circumstances where notification to a third party is required.
While information sharing is increasingly being recognized as one of the best ways to prevent cybersecurity incidents, it can create risks of its own. Some industries, particularly those which are highly regulated and tightly integrated, already engage is some form of lawful threat-sharing (e.g. financial sector). Other industries, especially those that are highly competitive, may be extremely reluctant to share information (e.g. retail sector).
In addition, Canadians in general (and an organization’s customers in particular) may take a dim view of an organization determining when it feels it is appropriate to share information with others, with no notice to those who may be affected by such information sharing. Conversely, organizations may be able to take refuge from the ire of customers or Canadians if there is, in the proposed regulations, a clear framework for when such sharing is required.
Question 19: Should the regulations set out specific circumstances where organizations would always be required to notify third parties of a data breach? If yes, in what circumstances?
- Record Keeping
The new record keeping provisions of the amendments have caused concern among organizations as the language is vague and, as drafted, does not appear to have much in the way of a threshold or trigger. Under Section 10.3(1) of PIPEDA, organizations that become aware of a breach of security safeguards must keep and maintain a record of the breach, regardless of the conclusion of their situational analysis into whether the breach poses a “real risk of significant harm”.
Organizations have been hoping that the regulations will provide clarity to what records will be required (e.g., individual incident reports for each suspected breach? Or will routine network logs suffice?). Depending upon how the regulations ultimately approach this issue, organizations may face extensive compliance projects in the future. Organizations which have aspired to a gold-standard in security safeguards and have gone beyond what is strictly necessary may find themselves re-thinking their approach as such additional security safeguards may impose increased record keeping obligations.
Not addressed by the government in its questions on the issue is the matter of privilege – a running record of day to day breaches of security safeguards will ultimately create a fertile evidentiary record for plaintiff’s counsel (and regulators).
The government has sought input on the following questions:
Question 20: Should the regulations list specific data fields for records or should it set-out a more flexible approach requiring “sufficient information to indicate the breach does not pose a real risk of significant harm” or similar?
Question 21: What information should the regulations require be included in a data breach record?
Question 22: Should the regulations specify a retention period for data breach records? If yes, how long would be considered a reasonable retention period?
Question 23: Should the regulations clarify that the individual(s) designated by the organization as those responsible for overseeing compliance with PIPEDA are those accountable for maintaining data breach records and providing them to the OPC upon request?
Question 24: Should the regulations clarify that a report made to the OPC satisfies the record-keeping requirement under section 10.3(1)?
Question 25: Should the regulations clarify that the obligation to maintain a data breach record applies only to data breaches for which the organization has actual knowledge?
Question 26: Should the regulations permit data breach records to take the form of periodic roll-ups that consolidate information concerning data breaches experienced by the organization over the applicable period? Or should the regulations specify that a separate record is required for each data breach experienced by the organization?
Organizations which handle personal information in the course of the commercial activities will want to monitor the regulation-making process closely and seriously consider providing their input, either alone or in conjunction with industry associations.