Teresa Scassa - Blog

This is Part III of a series of posts that look at the proposed amendments to Canada’s Artificial Intelligence and Data Act (which itself is still a Bill, currently before the INDU Committee for study). Part I provided a bit of context and a consideration of some of the new definitions in the Bill. Part II looked at the categories of ‘high-impact’ AI that the Bill now proposes to govern. This post looks at the changed role of the AI and Data Commissioner.

The original version of the Artificial Intelligence and Data Act (Part II of Bill C-27) received considerable criticism for its oversight mechanisms. Legal obligations for the ethical and transparent governance of AI, after all, depend upon appropriate oversight and enforcement for their effectiveness. Although AIDA proposed the creation of an AI and Data Commissioner (Commissioner), this was never meant to be an independent regulator. Ultimately, AIDA placed most of the oversight obligations in the hands of the Minister of Industry – the same Minister responsible for supporting the growth of Canada’s AI sector. Critics considered this to be a conflict of interest. A series of proposed amendments to AIDA are meant to address these concerns by reworking the role of the Commissioner.

Section 33(1) of AIDA makes it clear that the AI and Data Commissioner will be a “senior official of the department over which the Minister presides”, and their appointment involves being designated by the Minister. This has not changed, although the amendments would delete from this provision language stating that the Commissioner’s role is “to assist the Minister in the administration and enforcement” of AIDA.

The proposed amendments elevate the Commissioner somewhat, giving them a series of powers and duties, to which the Minister can add through delegation (s. 33(3)). So, for example, it will be the newly empowered Commissioner (Commissioner 2.0) who receives reports from those managing a general-purpose or high impact system where there are reasonable grounds to suspect that the use of the system has caused serious harm (s. 8.2(1)(e), s. 11(1)(g)). Commissioner 2.0 can also order someone managing or making available a general-purpose system to provide them with the accountability framework they are required to create under s. 12 (s. 13(1)) and can provide guidance or recommend corrections to that framework (s. 13(2)). Commissioner 2.0 can compel those making available or managing an AI system to provide the Commissioner with an assessment of whether the system is high impact, and in relation to which subclass of high impact systems set out in the schedule. Commissioner 2.0 can agree or disagree with the assessment, although if they disagree, their authority seems limited to informing the entity in writing with their reasons for disagreement.

More significant are Commissioner 2.0’s audit powers. Under the original version of AIDA, these were to be exercised by the Minister – the powers are now those of the Commissioner (s. 15(1)). Further, Commissioner 2.0 may order (previously this was framed as “require”) that the person either conduct an audit themselves or that the person engage the services of an independent auditor. The proposed amendments also empower the Commissioner to conduct an audit to determine if there is a possible contravention of AIDA. This strengthens the audit powers by ensuring that there is at least an option that is not at least somewhat under the control of the party being audited. The proposed amendments give Commissioner 2.0 additional powers necessary to conduct an audit and to carry out testing of an AI system (s. 15(2.1)). Where Commissioner 2.0 conducts an audit, they must provide the audited party with a copy of the report (s. 15(3.1)) and where the audit is conducted by the person responsible or someone retained by them, they must provide a copy to the Commissioner (s. 15(4)).

The Minister still retains some role with respect to audits. He or she may request that the Commissioner conduct an audit. In an attempt to preserve some independence of Commissioner 2.0, the Commissioner, when receiving such a request, may either carry out the audit or decline to do so on the basis that there are no reasonable grounds for an audit, so long as they provide the Minister with their reasons (s. 15.1(1)(b)). The Minister may also order a person to take actions to bring themselves into compliance with the law (s. 16) or to cease making available or terminate the operation of a system if the Minister considers compliance to be impossible (s. 16(b)) or has reasonable grounds to believe that the use of the system “gives rise to a risk of imminent and serious harm” (s. 17(1)).

As noted above, Commissioner 2.0 (a mere employee in the Minister’s department) will have order making powers under the amendments. This is something the Privacy Commissioner of Canada, an independent agent of Parliament, appointed by the Governor in Council, is hoping to get in Bill C-27. If so, it will be for the first time since the enactment of PIPEDA in 2000. Orders of Commissioner 2.0 or the Minister can become enforceable as orders of the Federal Court under s. 20.

Commissioner 2.0 is also empowered to share information with a list of federal or provincial government regulators where they have “reasonable grounds to believe that the information may be relevant to the administration or enforcement by the recipient of another Act of Parliament or of a provincial legislature.” (s. 26(1)). Reciprocally, under a new provision, federal regulators may also share information with the Commissioner (s. 26.1). Additionally, Commissioner 2.0 may “enter into arrangements” with different federal regulators and/or the Ministers of Health and Transport in order to assist those actors with the “exercise of their powers or the performance of their functions and duties” in relation to AI (s. 33.1). These new provisions strengthen a more horizontal, multi-regulator approach to governing AI which is an improvement in the Bill, although this might eventually need to be supplemented by corresponding legislative amendments – and additional funding – to better enable the other commissioners to address AI-related issues that fit within their areas of competence.

The amendments also impose upon Commissioner 2.0 a new duty to report on the administration and enforcement of AIDA – such a report is to be “published on a publicly available website”. (s. 35.1) The annual reporting requirement is important as it will increase transparency regarding the oversight and enforcement of AIDA. For his or her part, the Minister is empowered to publish information, where it is in the public interest, regarding any contravention of AIDA or where the use of a system gives rise to a serious risk of imminent harm (ss. 27 and 28).

Interestingly, AIDA, which provides for the potential imposition of administrative monetary penalties for contraventions of the Act does not indicate who is responsible for setting and imposing these penalties. Section 29(1)(g) makes it clear that “the persons or classes of persons who may exercise any power, or perform any duty or function, in relation to the [AMP] scheme” is left to be articulated in regulations.

The AIDA also makes it an offence under s. 30 for anyone to obstruct or provide false or misleading information to “the Minister, anyone acting on behalf of the Minister or an independent auditor in the exercise of their powers or performance of their duties or functions under this Part.” This remains unchanged from the original version of AIDA. Presumably, since Commissioner 2.0 would exercise a great many of the oversight functions, this is meant to apply to the obstruction or misleading of the Commissioner – but it will only do so if the Commissioner is characterized as someone “acting on behalf of the Minister”. This is not language of independence, but then there are other features of AIDA that also counter any view that even Commissioner 2.0 is truly independent (and I mean others besides the fact that they are an employee under the authority of the Minister and handpicked by the Minister). Most notable of these is that should the Commissioner become incapacitated or absent, or should they simply never be designated by the Minister, it is the Minister who will exercise their powers and duties (s. 33(4)).

In sum, then, the proposed amendments to AIDA attempt to give some separation between the Minister and Commissioner 2.0 in terms of oversight and enforcement. At the end of the day, however, Commissioner 2.0 is still the Minister’s hand-picked subordinate. Commissioner 2.0 does not serve for a specified term and has no security of tenure. In their absence, the Minister exercises their powers. It falls far short of independence.

My previous post looked at some of the new definitions in the proposed amendments to the Artificial Intelligence and Data Act (AIDA) which is Part III of Bill C-27. These include a definition of “high impact” AI, and a schedule of classes of high-impact AI (the Schedule is reproduced at the end of this post). The addition of the schedule changes AIDA considerably, and that is the focus of this post.

The first two classes in the Schedule capture contexts that can clearly affect individuals. Class 1 addresses AI used in most aspects of employment, and Class 2 relates to the provision of services. On the provision of services (which could include things like banking and insurance), the wording signals that it will apply to decision-making about the provision of services, their cost, or the prioritization of recipients. To be clear, AIDA does not prohibit systems with these functions. They are simply characterized as “high impact” so that they will be subject to governance obligations. A system to determine creditworthiness can still reject individuals; and companies can still prioritize preferred customers – as long as the systems are sufficiently transparent, free from bias and do not cause harm.

There is, however, one area which seems to fall through the cracks of Classes 1 & 2: rental accommodation. A lease is an interest in land – it is not a service. Human rights legislation in Canada typically refers to accommodation separately from services for this reason. AI applications are already being used to screen and select tenants for rental accommodation. In the midst of a housing crisis, this is surely an area that is high-impact and where the risks of harm from flawed AI to individuals and families searching for a place to live are significant. This gap needs to be addressed – perhaps simply by adding “or accommodation” after each use of the term “service” in Class 2.

Class 3 rightly identifies biometric systems as high risk. It also includes systems that use biometrics in “the assessment of an individual’s behaviour or state of mind.” Key to the scope of this section will be the definition of “biometric”. Some consider biometric data to be exclusively physiological data (fingerprints, iris scans, measurements of facial features, etc.). Yet others include behavioral data in this class if it is used for the second identified purpose – the assessment of behaviour or state of mind. Behavioural data, though, is potentially a very broad category. It can include data about a person’s gait, or their speech or keystroke patterns. Cast even more broadly, it could include things such as “geo-location and IP addresses”, “purchasing habits”, “patterns of device use” or even “browser history and cookies”. If that is the intention behind Class 3, then conventional biometric AI should be Part One of this class; Part Two should be the use of an AI system to assess an individual’s behaviour or state of mind (without referring specifically to biometrics in order to avoid confusion). This would also, importantly, capture the highly controversial area of AI for affect recognition. It would be unfortunate if the framing of the class as ‘biometrics’ led to an unduly narrow interpretation of the kind of systems or data involved. The explanatory note in the Minister’s cover letter for this provision seems to suggest (although it is not clear) that it is purely physiological biometric data that is intended for inclusion and not a broader category. If this is so, then Class 3 seems unduly narrow.

Class 4 is likely to be controversial. It addresses content moderation and the prioritization and presentation of content online and identifies these as high-impact algorithmic activities. Such systems are in widespread use in the online context. The explanatory note from the Minister observes that such systems “have important potential impacts on Canadians’ ability to express themselves, as well as pervasive effects at societal scale” (at p. 4). This is certainly true although the impact is less direct and obvious than the impact of a hiring algorithm, for example. Further, although an algorithm that presents a viewer of online streaming services with suggestions for content could have the effect of channeling a viewer’s attention in certain directions, it is hard to see this as “high impact” in many contexts, especially since there are multiple sources of suggestions for online viewing (including word of mouth). That does not mean that feedback loops and filter bubbles (especially in social media) do not contribute to significant social harms – but it does make this high impact class feel large and unwieldy. The Minister’s cover letter indicates that each of the high-impact classes presents “distinct risk profiles and consequently will require distinct risk management strategies.” (at p. 2). Further, he notes that the obligations that will be imposed “are intended to scale in proportion to the risks they present. A low risk use within a class would require correspondingly minimal mitigation effort.” (at p. 2). Much will clearly depend on regulations.

Class 5 relates to the use of AI in health care or emergency services, although it explicitly excludes medical devices because these are already addressed by Health Canada (which recently consulted on the regulation of AI-enabled medical devices). This category also demonstrates some of the complexity of regulating AI in Canada’s federal system. Many hospital-based AI technologies are being developed by researchers affiliated with the hospitals and who are not engaged in the interprovincial or international trade and commerce which is necessary for AIDA to apply. AIDA will only apply to those systems developed externally and in the context of international or interprovincial trade and commerce. While this will still capture many applications, it will not capture all – creating different levels of governance within the same health care context.

It is also not clear what is meant, in Class 5, by “use of AI in matters relating to health care”. This could be interpreted to mean health care that is provided within what is understood as the health care system. Understood more broadly, it could extend to health-related apps – for example, one of the many available AI-enabled sleep trackers, or an AI-enabled weight loss tool (to give just two examples). I suspect that what is intended is the former, even though, with health care in crisis and more people turning to alternate means to address their health issues, health-related AI technologies might well deserve to be categorized as high-impact.

Class 6 involves the use of an AI system by a court or administrative body “in making a determination in respect of an individual who is a party to proceedings before the court or administrative body.” In the first place, this is clearly not meant to apply to automated decision-making generally – it seems to be limited to judicial or quasi-judicial contexts. Class 6 must also be reconciled with s. 3 of AIDA, which provides that AIDA does not apply “with respect to a government institution as defined in s. 3 of the Privacy Act.” This includes the Immigration and Refugee Board, for example, as well as the Canadian Human Rights Commission, the Parole Board, and the Veterans Review and Appeal Board. Making sense of this, then, it would be the tools used by courts or tribunals and developed or deployed in the course of interprovincial or international trade and commerce that would be considered high impact. The example given in the Minister’s letter seems to support this – it is of an AI system that provides an assessment of “risk of recidivism based on historical data” (at p. 5).

However, Class 6 is confusing because it identifies the context rather than the tools as high impact. Note that the previous classes address the use of AI “in matters relating to” the subject matter of the class, whereas class 6 identifies actors – the use of AI by a court or tribunal. There is a different focus. Yet the same tools used by courts and tribunals might also be used by administrative bodies or agencies that do not hold hearings or that are otherwise excluded from the application of AIDA. For example, in Ewert v. Canada, the Supreme Court of Canada considered an appeal by a Métis man who challenged the use of recidivism-risk assessment tools by Correctional Services of Canada (to which AIDA would not apply according to s. 3). If this type of tool is high-risk, it is so whether it is used by Correctional Services or a court. This suggests that the framing of Class 6 needs some work. It should perhaps be reworded to identify tools or systems as high impact if they are used to determine the rights, entitlements or status of individuals.

Class 7 addresses the use of an AI system to assist a peace officer “in the exercise and performance of their law enforcement powers, duties and function”. Although “peace officer” receives the very broad interpretation found in the Criminal Code, that definition is modified in the AIDA by language that refers to the exercise of specific law enforcement powers. This should still capture the use of a broad range of AI-enabled tools and technologies. It is an interesting question whether AIDA might apply more fulsomely to this class of AI systems (not just those developed in the course of interprovincial or international trade) as it might be considered to be rooted in the federal criminal law power.

AIDA Schedule:

High-Impact Systems — Uses

1. The use of an artificial intelligence system in matters relating to determinations in respect of employment, including recruitment, referral, hiring, remuneration, promotion, training, apprenticeship, transfer or termination.

2. The use of an artificial intelligence system in matters relating to

(a) the determination of whether to provide services to an individual;

(b) the determination of the type or cost of services to be provided to an individual; or

(c) the prioritization of the services to be provided to individuals.

3. The use of an artificial intelligence system to process biometric information in matters relating to

(a) the identification of an individual, other than in cases in which the biometric information is processed with the individual’s consent to authenticate their identity; or

(b) the assessment of an individual’s behaviour or state of mind.

4. The use of an artificial intelligence system in matters relating to

(a) the moderation of content that is found on an online communications platform, including a search engine or social media service; or

(b) the prioritization of the presentation of such content.

5. The use of an artificial intelligence system in matters relating to health care or emergency services, excluding a use referred to in any of paragraphs (a) to (e) of the definition device in section 2 of the Food and Drugs Act that is in relation to humans.

6. The use of an artificial intelligence system by a court or administrative body in making a determination in respect of an individual who is a party to proceedings before the court or administrative body.

7. The use of an artificial intelligence system to assist a peace officer, as defined in section 2 of the Criminal Code, in the exercise and performance of their law enforcement powers, duties and functions.

Note: This is the first in a series of posts that will look at the proposed amendments to Canada's Artificial Intelligence and Data Act, which is Part III of Bill C-27, currently before Parliament. The amendments are extensive and have only just been introduced. Please consider these assessments to be preliminary.

Canada’s Artificial Intelligence and Data Act (AIDA) (Part III of Bill C-27) has passed second reading and is currently before the INDU Committee for study. Early in this committee process, the Minister of Industry Philippe Champagne announced that his department was working on amendments to AIDA in response to considerable criticism. Those amendments have now been tabled for consideration by the committee.

One of the criticisms of the Bill was that it left almost all of its substance to be developed in regulations. It is unsurprising, then, that the amendments are almost as long as the original bill. While it is certainly the case that the amendments contain more detail than the original text, some of the additional length is attributable to new provisions intended to address generative AI systems. This highlights just how quickly things are moving in the AI space, as generative AI was not on anyone’s legislative radar when Bill C-27 was introduced in June 2022.

One of the criticisms of AIDA was the absence of any specific prior consultation before its appearance in Bill C-27. This, combined with its lack of substance on many issues, raised basic concerns about how it would apply and to what. For example, AIDA was to govern “high-impact” AI systems, but the definition of such systems was left to regulations. Concerns were also raised about oversight being largely in the hands of the Minister of Industry who is also responsible for supporting Canada’s AI sector.

The proposed amendments demonstrate that ISED has been listening to the feedback it has received since June 2022, just as it has been adapting to the challenges of generative AI, and engaging with its international partners on AI governance issues. The amendments, which include new definitions, more explicit obligations, and governance principles for generative AI, will make AIDA a better bill. They may be enough to garner sufficient support to pass it into law, something which the Minister describes as “pivotal”.

This is the first in a series of posts that will explore some of the changes proposed to AIDA – as well as some of the remaining issues. This post addresses some of the new definitions.

The proposed AIDA amendments propose a new definition of “artificial intelligence system” which would read: “a technological system that, using a model, makes inferences in order to generate output, including predictions, recommendations or decisions” (s. 2). This provides greater alignment with the OECD definition of an AI system (“An AI system is a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments. Different AI systems vary in their levels of autonomy and adaptiveness after deployment.”) It is an improvement over the previous definition, which was criticized for being too specific about the types of techniques used in AI. It is unclear, though, why the new AIDA definition does not include “content” as an output as is the case with the OECD definition. The AIDA definition is also supplemented by a separate definition for a “general-purpose system”, which is “an artificial intelligence system that is designed to be adapted for use, in many fields and for many purposes and activities, including fields, purposes and activities not contemplated during the system’s development” (s. 5(1)). There is a further definition for a “machine learning model”, which is “a digital representation of patterns identified in data through the automated processing of the data using an algorithm designed to enable the recognition or replication of those patterns”. A new s. 5(2) makes it clear that the definition of AI system includes general-purpose systems, and that general-purpose systems can also be high-impact. These new definitions reflect the major changes in both the technology and in the evolving regulatory context in the short time since AIDA was introduced. They also shape a new framework for obligations under the legislation.

The proposed amendments also contain a definition of “high-impact system”: “an artificial intelligence system of which at least one of the intended uses may reasonably be concluded to fall within a class of uses set out in the schedule”. (s. 5(1)). The previous version of AIDA left the articulation of “high impact” to future regulations. The schedule sets out a list of classes that describe certain uses. These are:

High-Impact Systems — Uses

1. The use of an artificial intelligence system in matters relating to determinations in respect of employment, including recruitment, referral, hiring, remuneration, promotion, training, apprenticeship, transfer or termination.

2. The use of an artificial intelligence system in matters relating to

(a) the determination of whether to provide services to an individual;

(b) the determination of the type or cost of services to be provided to an individual; or

(c) the prioritization of the services to be provided to individuals.

3. The use of an artificial intelligence system to process biometric information in matters relating to

(a) the identification of an individual, other than in cases in which the biometric information is processed with the individual’s consent to authenticate their identity; or

(b) the assessment of an individual’s behaviour or state of mind.

4. The use of an artificial intelligence system in matters relating to

(a) the moderation of content that is found on an online communications platform, including a search engine or social media service; or

(b) the prioritization of the presentation of such content.

5. The use of an artificial intelligence system in matters relating to health care or emergency services, excluding a use referred to in any of paragraphs (a) to (e) of the definition device in section 2 of the Food and Drugs Act that is in relation to humans.

6. The use of an artificial intelligence system by a court or administrative body in making a determination in respect of an individual who is a party to proceedings before the court or administrative body.

7. The use of an artificial intelligence system to assist a peace officer, as defined in section 2 of the Criminal Code, in the exercise and performance of their law enforcement powers, duties and functions.

(Note: the classes in this schedule will be the subject of the next blog post)

The list is not intended to be either closed or permanent. Under a proposed s. 36.1, the Governor in Council (GinC) can enact regulations amending the schedule by adding, modifying, or deleting a category of use. Any such decision by the GinC is to be guided by criteria set out in s. 36.1. These include the risks of adverse impact on “the economy or any other aspect of Canadian society and on individuals, including on individual’s health and safety and on their rights recognized in international human rights treaties to which Canada is a party”. The GinC must also consider the “severity and extent” of any adverse impacts, as well as the “social and economic circumstances of any individuals who may experience” such impacts. A final consideration is whether the uses in the category are adequately addressed under another Act of Parliament or of a provincial legislature.

The AIDA only applies to “high impact” systems, and since there is no screening or registration process, it is up to those who manage or make such systems available to identify them as such and to meet the obligations set out in the law. A proposed s. 14 would empower the AI and Data Commissioner to order a person who makes available or who manages an AI system to provide the Commissioner with their assessment of whether the system is a high impact system, a general purpose system (which can also be high impact), or a machine learning model intended to be incorporated into a high impact system.

My next post will look at the classes of “high-impact” AI as set out in the Schedule.

On October 26, 2023, I appeared as a witness before the INDU Committee of the House of Commons which is holding hearings on Bill C-27. Although I would have preferred to address the Artificial Intelligence and Data Act, it was clear that the Committee was prioritizing study of the Consumer Protection and Privacy Act in part because the Minister of Industry had yet to produce the text of amendments to the AI and Data Act which he had previously outlined in a letter to the Committee Chair. It is my understanding that witnesses will not be called twice. As a result, I will be posting my comments on the AI and Data Act on my blog.

The other witnesses heard at the same time included Colin Bennett, Michael Geist, Vivek Krishnamurthy and Brenda McPhail. The recording of that session is available here.

__________

Thank you, Mr Chair, for the invitation to address this committee.

I am a law professor at the University of Ottawa, where I hold the Canada Research Chair in Information Law and Policy. I appear today in my personal capacity. I have concerns with both the CPPA and AIDA. Many of these have been communicated in my own writings and in the report submitted to this committee by the Centre for Digital Rights. My comments today focus on the Consumer Privacy Protection Act. I note, however, that I have very substantial concerns about the AI and Data Act and would be happy to answer questions on it as well.

Let me begin by stating that I am generally supportive of the recommendations of Commissioner Dufresne for the amendment of Bill C-27 set out in his letter of April 26, 2023, to the Chair of this Committee. I will also address 3 other points.

The Minister has chosen to retain consent as the backbone of the CPPA, with specific exceptions to consent. One of the most significant of these is the “legitimate interest” exception in s. 18(3). This allows organizations to collect or use personal information without knowledge or consent if it is for an activity in which an organization has a legitimate interest. There are guardrails: the interest must outweigh any adverse effects on the individual; it must be one which a reasonable person would expect; and the information must not be collected or used to influence the behaviour or decisions of the individual. There are also additional documentation and mitigation requirements.

The problem lies in the continuing presence of “implied consent” in section 15(5) of the CPPA. PIPEDA allowed for implied consent because there were circumstances where it made sense, and there was no “legitimate interest” exception. However, in the CPPA, the legitimate interest exception does the work of implied consent. Leaving implied consent in the legislation provides a way to get around the guardrails in s. 18(3) (an organization can opt for the ‘implied consent’ route instead of legitimate interest). It will create confusion for organizations that might struggle to understand which is the appropriate approach. The solution is simple: get rid of implied consent. I note that “implied consent” is not a basis for processing under the GDPR. Consent must be express or processing must fall under another permitted ground.

My second point relates to s. 39 of the CPPA, which is an exception to an individual’s knowledge and consent where information is disclosed to a potentially very broad range of entities for “socially beneficial purposes”. Such information need only be de-identified – not anonymized – making it more vulnerable to reidentification. I question whether there is social licence for sharing de-identified rather than anonymized data for these purposes. I note that s. 39 was carried over verbatim from C-11, when “de-identify” was defined to mean what we understand as “anonymize”.

Permitting disclosure for socially beneficial purposes is a useful idea, but s. 39, especially with the shift in meaning of “de-identify”, lacks necessary safeguards. First, there is no obvious transparency requirement. If we are to learn anything from the ETHI Committee inquiry into PHAC’s use of Canadians’ mobility data, transparency is fundamentally important. At the very least, there should be a requirement that written notice of data sharing for socially beneficial purposes be given to the Privacy Commissioner of Canada; ideally there should also be a requirement for public notice. Further, s. 39 should provide that any such sharing be subject to a data sharing agreement, which should also be provided to the Privacy Commissioner. None of this is too much to ask where Canadians’ data are conscripted for public purposes. Failure to ensure transparency and some basic measure of oversight will undermine trust and legitimacy.

My third point relates to the exception to knowledge and consent for publicly available personal information. Bill C-27 reproduces PIPEDA’s provision on publicly available personal information, providing in s. 51 that “An organization may collect, use or disclose an individual’s personal information without their knowledge or consent if the personal information is publicly available and is specified by the regulations.” We have seen the consequences of data scraping from social media platforms in the case of Clearview AI, which used scraped photographs to build a massive facial recognition database. The Privacy Commissioner takes the position that personal information on social media platforms does not fall within the “publicly available personal information” exception. Yet not only could this approach be upended in the future by the new Personal Information and Data Protection Tribunal, it could also easily be modified by new regulations. Recognizing the importance of s. 51, former Commissioner Therrien had recommended amending it to add that the publicly available personal information be such “that the individual would have no reasonable expectation of privacy”. An alternative is to incorporate the text of the current Regulations Specifying Publicly Available Information into the CPPA, revising them to clarify scope and application in our current data environment. I would be happy to provide some sample language.

This issue should not be left to regulations. The amount of publicly available personal information online is staggering, and it is easily susceptible to scraping and misuse. It should be clear and explicit in the law that personal data cannot be harvested from the internet, except in limited circumstances set out in the statute.

Finally, I add my voice to those of so many others in saying that the data protection obligations set out in the CPPA should apply to political parties. It is unacceptable that they do not.

The following is a short excerpt from a new paper which looks at the public sector use of private sector personal data (Teresa Scassa, “Public Sector Use of Private Sector Personal Data: Towards Best Practices”, forthcoming in (2024) 47:2 Dalhousie Law Journal ) The full pre-print version of the paper is available here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4538632

Governments seeking to make data-driven decisions require the data to do so. Although they may already hold large stores of administrative data, their ability to collect new or different data is limited both by law and by practicality. In our networked, Internet of Things society, the private sector has become a source of abundant data about almost anything – but particularly about people and their activities. Private sector companies collect a wide variety of personal data, often in high volumes, rich in detail, and continuously over time. Location and mobility data, for example, are collected by many different actors, from cellular service providers to app developers. Financial sector organizations amass rich data about the spending and borrowing habits of consumers. Even genetic data is collected by private sector companies. The range of available data is constantly broadening as more and more is harvested, and as companies seek secondary markets for the data they collect.

Public sector use of private sector data is fraught with important legal and public policy considerations. Chief among these is privacy since access to such data raises concerns about undue government intrusion into private lives and habits. Data protection issues implicate both public and private sector actors in this context, and include notice and consent, as well as data security. And, where private sector data is used to shape government policies and actions, important questions about ethics, data quality, the potential for discrimination, and broader human rights questions also arise. Alongside these issues are interwoven concerns about transparency, as well as necessity and proportionality when it comes to the conscription by the public sector of data collected by private companies.

This paper explores issues raised by public sector access to and use of personal data held by the private sector. It considers how such data sharing is legally enabled and within what parameters. Given that laws governing data sharing may not always keep pace with data needs and public concerns, this paper also takes a normative approach which examines whether and in what circumstances such data sharing should take place. To provide a factual context for discussion of the issues, the analysis in this paper is framed around two recent examples from Canada that involved actual or attempted access by government agencies to private sector personal data for public purposes. The cases chosen are different in nature and scope. The first is the attempted acquisition and use by Canada’s national statistics organization, Statistics Canada (StatCan), of data held by credit monitoring companies and financial institutions to generate economic statistics. The second is the use, during the COVID-19 pandemic, of mobility data by the Public Health Agency of Canada (PHAC) to assess the effectiveness of public health policies in reducing the transmission of COVID-19 during lockdowns. The StatCan example involves the compelled sharing of personal data by private sector actors; while the PHAC example involves a government agency that contracted for the use of anonymized data and analytics supplied by private sector companies. Each of these instances generated significant public outcry. This negative publicity no doubt exceeded what either agency anticipated. Both believed that they had a legal basis to gather and/or use the data or analytics, and both believed that their actions served the public good. Yet the outcry is indicative of underlying concerns that had not properly been addressed.

Using these two quite different cases as illustrations, the paper examines the issues raised by the use of private sector data by government. Recognizing that such practices are likely to multiply, it also makes recommendations for best practices. Although the examples considered are Canadian and are shaped by the Canadian legal context, most of the issues they raise are of broader relevance. Part I of this paper sets out the two case studies that are used to tease out and illustrate the issues raised by public sector use of private sector data. Part II discusses the different issues and makes recommendations.

The full pre-print version of the paper is available here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4538632

A recent decision of the Federal Court of Canada ends (subject to any appeal) the federal Privacy Commissioner’s attempt to obtain an order against Facebook in relation to personal information practices linked to the Cambridge Analytica scandal. Following a joint investigation with British Columbia’s Information and Privacy Commissioner, the Commissioners had issued a Report of Findings in 2019. The Report concluded that Facebook had breached Canada’s Personal Information Protection and Electronic Documents Act (PIPEDA) and B.C.’s Personal Information Protection Act by failing to obtain appropriate consent, failing to adequately safeguard the data of its users and failing to be accountable for the data under its control. Under PIPEDA, the Privacy Commissioner has no order-making powers and can only make non-binding recommendations. For an order to be issued under PIPEDA, an application must be made to the Federal Court under s. 15, either by the complainant, or by the Privacy Commissioner with the complainant’s permission. The proceeding before the court is de novo, meaning that the court renders its own decision on whether there has been a breach of PIPEDA based upon the evidence presented to it.

The Cambridge Analytica scandal involved a researcher who developed a Facebook app. Through this app, the developer collected user data, ostensibly for research purposes. That data was later disclosed to third parties who used it to develop “psychographic” models for purposes of targeting political messages towards segments of Facebook users” (at para 35). It is important to note here that the complaint was not against the app developer, but rather against Facebook. Essentially, the complainants were concerned that Facebook did not adequately protect its users’ privacy. Although it had put in place policies and requirements for third party app developers, the complainants were concerned that it did not adequately monitor the third-party compliance with its policies.

The Federal Court dismissed the Privacy Commissioner’s application largely because of a lack of evidence to establish that Facebook had failed to meet its PIPEDA obligations to safeguard its users’ personal information. Referring to it as an “evidentiary vacuum” (para 71), Justice Manson found that there was a lack of expert evidence regarding what Facebook might have done differently. He also found that there was no evidence from users regarding their expectations of privacy on Facebook. The Court chastised the Commissioner, stating “ultimately it is the Commissioner’s burden to establish a breach of PIPEDA on the basis of evidence, not speculation and inferences derived from a paucity of material facts” (at para 72). Justice Manson found the evidence presented by the Commissioner to be unpersuasive, speculative, and required the court to draw “unsupported inferences”. He was unsympathetic to the Commissioner’s explanation that it did not use its statutory powers to compel evidence (under s. 12.1 of PIPEDA) because “Facebook would not have complied or would have had nothing to offer” (at para 72). Justice Manson noted that had Facebook failed to comply with requests under s. 12.1, the Commissioner could have challenged the refusal.

Yet there is more to this decision than just a dressing down of the Commissioner’s approach to the case. In discussing “meaningful consent” under PIPEDA, Justice Manson frames the question before the court as “whether Facebook made reasonable efforts to ensure users and users’ Facebook friends were advised of the purposes for which their information would be used by third-party applications” (at para 63). This argument is reflected in the Commissioner’s position that Facebook should have done more to ensure that third party app developers on its site complied with their contractual obligations, including those that required developers to obtain consent from app users to the collection of personal data. Facebook’s position was that PIPEDA only requires that it make reasonable efforts to protect the personal data of its users, and that it had done so through its “combination of network-wide policies, user controls and educational resources” (at para 68). It is here that Justice Manson emphasizes the lack of evidence before him, noting that it is not clear what else Facebook could have reasonably been expected to do. In making this point, he states:

There is no expert evidence as to what Facebook could feasibly do differently, nor is there any subjective evidence from Facebook users about their expectations of privacy or evidence that any user did not appreciate the privacy issues at stake when using Facebook. While such evidence may not be strictly necessary, it would have certainly enabled the Court to better assess the reasonableness of meaningful consent in an area where the standard for reasonableness and user expectations may be especially context dependent and ever-evolving. (at para 71) [My emphasis].

This passage should be deeply troubling to those concerned about privacy. By referring to the reasonable expectation of privacy in terms of what users might expect in an ever-evolving technological context, Justice Manson appears to abandon the normative dimensions of the concept. His comments lead towards a conclusion that the reasonable expectation of privacy is an ever-diminishing benchmark as it becomes increasingly naïve to expect any sort of privacy in a data-hungry surveillance society. Yet this is not the case. The concept of the “reasonable expectation of privacy” has significant normative dimensions, as the Supreme Court of Canada reminds us in R. v. Tessling and in the case law that follows it. In Tessling, Justice Binnie noted that subjective expectations of privacy should not be used to undermine the privacy protections in s. 8 of the Charter, stating that “[e]xpectation of privacy is a normative rather than a descriptive standard.” Although this comment is made in relation to the Charter, a reasonable expectation of privacy that is based upon the constant and deliberate erosion of privacy would be equally meaningless in data protection law. Although Justice Manson’s comments about the expectation of privacy may not have affected the outcome of this case, they are troublesome in that they might be picked up by subsequent courts or by the Personal Information and Data Protection Tribunal proposed in Bill C-27.

The decision also contains at least two observations that should set off alarm bells with respect to Bill C-27, a bill to reform PIPEDA. Justice Manson engages in some discussion of the duty of an organization to safeguard information that it has disclosed to a third party. He finds that PIPEDA imposes obligations on organizations with respect to information in their possession, and information transferred for processing. In the case of prospective business transactions, an organization sharing information with a potential purchaser must enter into an agreement to protect that information. However, Justice Manson interprets this specific reference to a requirement for such an agreement to mean that “[i]f an organization were required to protect information transferred to third parties more generally under the safeguarding principle, this provision would be unnecessary” (at para 88). In Bill C-27, s. 39, for example, permits organizations to share de-identified (not anonymized) personal information with certain third parties without the knowledge or consent of individuals for ‘socially beneficial’ purposes without imposing any requirement to put in place contractual provisions to safeguard that information. The comments of Justice Manson clearly highlight the deficiencies of s. 39 which must be amended to include a requirement for such safeguards.

A second issue relates to the human-rights based approach to privacy which both the former Privacy Commissioner Daniel Therrien and the current Commissioner Philippe Dufresne have openly supported. Justice Manson acknowledges, that the Supreme Court of Canada has recognized the quasi-constitutional nature of data protection laws such as PIPEDA, because “the ability of individuals to control their personal information is intimately connected to their individual autonomy, dignity, and privacy” (at para 51). However, neither PIPEDA nor Bill C-27 take a human-rights based approach. Rather, they place personal and commercial interests in personal data on the same footing. Justice Manson states: “Ultimately, given the purpose of PIPEDA is to strike a balance between two competing interests, the Court must interpret it in a flexible, common sense and pragmatic manner” (at para 52). The government has made rather general references to privacy rights in the preamble of Bill C-27 (though not in any preamble to the proposed Consumer Privacy Protection Act) but has steadfastly refused to reference the broader human rights context of privacy in the text of the Bill itself. We are left with a purpose clause that acknowledges “the right of privacy of individuals with respect to their personal information” in a context in which “significant economic activity relies on the analysis, circulation and exchange of personal information”. The purpose clause finishes with a reference to the need of organizations to “collect, use or disclose personal information for purposes that a reasonable person would consider appropriate in the circumstances.” While this reference to the “reasonable person” should highlight the need for a normative approach to reasonable expectations as discussed above, the interpretive approach adopted by Justice Manson also makes clear the consequences of not adopting an explicit human-rights based approach. Privacy is thrown into a balance with commercial interests without fundamental human rights to provide a firm backstop.

Justice Manson seems to suggests that the Commissioner’s approach in this case may flow from frustration with the limits of PIPEDA. He describes the Commissioner’s submissions as “thoughtful pleas for well-thought-out and balanced legislation from Parliament that tackles the challenges raised by social media companies and the digital sharing of personal information, not an unprincipled interpretation from this Court of existing legislation that applies equally to a social media giant as it may apply to the local bank or car dealership.” (at para 90) They say that bad cases make bad law; but bad law might also make bad cases. The challenge is to ensure that Bill C-27 does not reproduce or amplify deficiencies in PIPEDA.

The government of the United Kingdom has published a consultation paper seeking input into its proposal for AI regulation. The paper is aptly titled A pro-innovation approach to AI regulation, since it restates that point insistently throughout the document. The UK proposal provides an interesting contrast to Canada’s AI governance bill currently before Parliament.

Both Canada and the UK set out to regulate AI systems with the twin goals of supporting innovation on the one hand, and building trust in AI on the other. (Note here that the second goal is to build trust in AI, not to protect the public. Although the protection of the public is acknowledged as one way to build trust, there is a subtle distinction here). However, beyond these shared goals, the proposals are quite different. Canada’s approach in Part 3 of Bill C-27 (the Artificial Intelligence and Data Act (AIDA)) is to create a framework to regulate as yet undefined “high impact” AI. The definition of “high impact” as well as many other essential elements of the bill are left to be articulated in regulations. According to a recently published companion document to the AIDA, leaving so much of the detail to regulations is how the government proposes to keep the law ‘agile’ – i.e. capable of responding to a rapidly evolving technological context. The proposal would also provide some governance for anonymized data by imposing general requirements to document the use of anonymized personal information in AI innovation. The Minister of Innovation is made generally responsible for oversight and enforcement. For example, the AIDA gives the Minister of Innovation the authority (eventually) to impose stiff administrative monetary penalties on bad actors. The Canadian approach is similar to that in the EU AI Act in that it aims for a broad regulation of AI technologies, and it chooses legislation as the vehicle to do so. It is different in that the EU AI Act is far more detailed and prescriptive; the AIDA leaves the bulk of its actual legal requirements to be developed in regulations.

The UK proposal is notably different from either of these approaches. Rather than create a new piece of legislation and/or a new regulatory authority, the UK proposes to set out five principles for responsible AI development and use. Existing regulators will be encouraged and, if necessary, specifically empowered, to regulate AI according to these principles within their spheres of regulatory authority. Examples of regulators who will be engaged in this framework include the Information Commissioner’s Office, regulators for human rights, consumer protection, health care products and medical devices, and competition law. The UK scheme also accepts that there may need to be an entity within government that can perform some centralized support functions. These may include monitoring and evaluation, education and awareness, international interoperability, horizon scanning and gap analysis, and supporting testbeds and sandboxes. Because of the risk that some AI technologies or issues may fall through the cracks between existing regulatory schemes, the government anticipates that regulators will assist government in identifying gaps and proposing appropriate actions. These could include adapting the mandates of existing regulators or providing new legislative measures if necessary.

Although Canada’s federal government has labelled its approach to AI regulation as ‘agile’, it is clear that the UK approach is much closer to the concept of agile regulation. Encouraging existing regulators to adapt the stated AI principles to their remit and to provide guidance on how they will actualize these principles will allow them to move quickly, so long as there are no obvious gaps in legal authority. By contrast, even once passed, it will take at least two years for Canada’s AIDA to have its normative blanks filled in by regulations. And, even if regulations might be somewhat easier to update than statutes, guidance is even more responsive, giving regulators greater room to manoeuvre in a changing technological landscape. Embracing the precepts of agile regulation, the UK scheme emphasizes the need to gather data about the successes and failures of regulation itself in order to adapt as required. On the other hand, while empowering (and resourcing) existing regulators will have clear benefits in terms of agility, the regulatory gaps could well be important ones – with the governance of large language models such as ChatGPT as one example. While privacy regulators are beginning to flex their regulatory muscles in the direction of ChatGPT, data protection law will only address a subset of the issues raised by this rapidly evolving technology. In Canada, AIDA’s governance requirements will be specific to risk-based regulation of AI, and will apply to all those who design, develop or make AI systems available for use (unless of course they are explicitly excluded under one of the many actual and potential exceptions).

Of course, the scheme in the AIDA may end up as more of a hybrid between the EU and the UK approaches in that the definition of “high impact” AI (to which the AIDA will apply) may be shaped not just by the degree of impact of the AI system at issue but also by the existence of other suitable regulatory frameworks. In other words, the companion document suggests that some existing regulators (health, consumer protection, human rights, financial institutions) have already taken steps to extend their remit to address the use of AI technologies within their spheres of competence. In this regard, the companion document speaks of “regulatory gaps that must be filled” by a statute such as AIDA as well as the need for the AIDA to integrate “seamlessly with existing Canadian legal frameworks”. Although it is still unclear whether the AIDA will serve only to fill regulatory gaps, or will provide two distinct layers of regulation in some cases, one of the criteria for identifying what constitutes a “high impact” system includes “[t]he degree to which the risks are adequately regulated under another law”. The lack of clarity in the Canadian approach is one of its flaws.

There is a certain attractiveness in the idea of a regulatory approach like that proposed by the UK – one that begins with existing regulators being both specifically directed and further enabled to address AI regulation within their areas of responsibility. As noted earlier, it seems far more agile than Canada’s rather clunky bill. Yet such an approach is much easier to adopt in a unitary state than in a federal system such as Canada’s. In Canada, some of the regulatory gaps are with respect to matters otherwise under provincial jurisdiction. Thus, it is not so simple in Canada to propose to empower and resource all implicated regulators, nor is it as easy to fill gaps once they are identified. These regulators and the gaps between them might fall under the jurisdiction of any one of 13 different governments. The UK acknowledges (and defers) its own challenges in this regard with respect to devolution at paragraph 113 of its white paper, where it states: “We will continue to consider any devolution impacts of AI regulation as the policy develops and in advance of any legislative action”. Instead, the AIDA, Canada leverages its general trade and commerce power in an attempt to provide AI governance that is as comprehensive as possible. It isn’t pretty (since it will not capture all AI innovation that might have impacts on people) but it is part of the reality of the federal state (or the state of federalism) in which we find ourselves.

The federal government’s proposed Artificial Intelligence and Data Act (AIDA) is currently before Parliament as part of Bill C-27, a bill that will also reform Canada’s private sector data protection law. The AIDA, which I have discussed in more detail in a series of blog posts (here, here, and here), has been criticized for being a shell of a law with essential components (including the definition of the “high impact AI” to which it will apply) being left to as-yet undrafted regulations. The paucity of detail in the AIDA, combined with the lack of public consultation, has prompted considerable frustration and concern from AI developers and from civil society alike. In response to these concerns, the government published, on March 13, 2023, a companion document that explains the government’s thinking behind the AIDA. The document is a useful read as it makes clear some of the rationales for different choices that have been made in the bill. It also obliquely engages with many of the critiques that have been leveled at the AIDA. Unlike a consultation document, however, where feedback is invited to improve what is being proposed, the companion document is essentially an apology (in the Greek sense of the word) – something that is written in defense or explanation. At this stage, any changes will have to come as amendments to the bill.

Calling this a ‘companion document’ also somewhat tests the notion of “companion”, since it was published nine months after the AIDA was introduced in Parliament in June 2022. The document explains that the government seeks to take “the first step towards a new regulatory system designed to guide AI innovation in a positive direction, and to encourage the responsible adoption of AI technologies by Canadians and Canadian businesses.” The AIDA comes on the heels of the European Union’s draft AI Act – a document that is both more comprehensive and far more widely consulted upon. Pressure on Canada to regulate AI is heightened by the activity in the EU. This is evident in the introduction to the companion document, which speaks of the need to work with international partners to achieve global protection for Canadians and to ensure that “Canadian firms can be recognized internationally as meeting robust standards.”

An important critique of the AIDA has been that it will apply only to “high impact” AI. By contrast, the EU AI Act sets a sliding scale of obligations, with the most stringent obligations applying to high risk applications, and minimal obligations for low risk AI. In the AIDA companion document, there is no explanation of why the AIDA is limited to high impact AI. The government explains that defining the scope of the Act in regulations will allow for greater precision, as well as for updates as technology progresses. The companion document offers some clues about what the government considers relevant to determining whether an AI system is high-impact. Factors include the type of harm, the severity of harm, and the scale of use. Although this may help understand the concept of high impact, it does not explain why governance was only considered for high and not medium or low impact AI. This is something that cannot be fixed by the drafting of regulations. The bill would have to be specifically amended to provide for governance for AI with different levels of impact according to a sliding scale of obligations.

Another important critique of the AIDA has been that it unduly focuses on individual rather than collective or broader harms. As the US’s NIST AI Risk Management Framework aptly notes, AI technologies “pose risks that can negatively impact individuals, groups, organizations, communities, society, the environment and the planet” (at p. 1). The AIDA companion document addresses this critique by noting that the bill is concerned both with individual harms and with systemic bias (defined as discrimination). Yet, while it is crucially important to address the potential for systemic bias in AI, this is not the only collective harm that should be considered. The potential for AI to be used to generate and spread disinformation or misinformation, for example, can create a different kind of collective harm. Flawed AI could potentially also result in environmental damage that is the concern of all. The companion document does little to address a broader notion of harm – but how can it? The AIDA specifically refers to and defines “individual harm”, and also addresses biased output as discriminatory within the meaning of the Canadian Human Rights Act. Only amendments to the bill can broaden its scope to encompass other forms of collective harm. Such amendments are essential.

Another critique of the AIDA is that it relies for its oversight on the same Ministry that is responsible for promoting and supporting AI innovation in Canada. The companion document tackles this concern, citing the uniqueness of the AI context, and stating that “administration and enforcement decisions have important implications for policy”, such that oversight and the encouragement of innovation “would need to be [sic] work in close collaboration in the early years of the framework under the direction of the Minister.” The Minister will be assisted by a Ministry staffer who will be designated the AI and Data Commissioner. The document notes that the focus in the early days of the legislation will be on helping organizations become compliant: “The Government intends to allow ample time for the ecosystem to adjust to the new framework before enforcement actions are undertaken.” The ample time will include the (at least) two years before the necessary regulations are drafted (though note that if some key regulations are not drafted, the law will never take effect), as well as any subsequent ‘adjustment’ time. Beyond this, the document is quite explicit that compliance and enforcement should not get unnecessarily in the way of the industry. The AIDA contains other mechanisms, including requiring companies to hire their own auditors for audits and having an appointed Ministerial advisory committee to reassure those who remain concerned about governance. Yet these measures do nothing to address a core lack of independent oversight. This lack is particularly noteworthy given that the same government has proposed the creation of an ill-advised Personal Information and Data Protection Tribunal (in Part II of Bill C-27) in order to establish another layer between the Privacy Commissioner and the enforcement of Bill C-27’s proposed Consumer Privacy Protection Act. It is difficult to reconcile the almost paranoid approach taken to the Privacy Commissioner’s role with the in-house, “we’re all friends here” approach to AI governance in the AIDA. It is hard to see how this lack of a genuine oversight framework can be fixed without a substantial rewrite of the bill.

And that brings us to the reality that we must confront with this bill: AI technologies are rapidly advancing and are already having significant impacts on our lives. The AIDA is deeply flawed, and the lack of consultation is profoundly disturbing. Yet, given the scarcity of space on Parliament’s agenda and the generally fickle nature of politics, the failure of the AIDA could lead to an abandonment of attempts to regulate in this space – or could very substantially delay them. As debate unfolds over the AIDA, Parliamentarians will have to ask themselves the unfortunate question of whether the AIDA is unsalvageable, or whether it can be sufficiently amended to be better than no law at all.

A recent decision of the Federal Court of Canada exposes the tensions between access to information and privacy in our data society. It also provides important insights into how reidentification risk should be assessed when government agencies or departments respond to requests for datasets with the potential to reveal personal information.

The case involved a challenge by two journalists to Health Canada’s refusal to disclose certain data elements in a dataset of persons permitted to grow medical marijuana for personal use under the licensing scheme that existed before the legalization of cannabis. [See journalist Molly Hayes’ report on the story here]. Health Canada had agreed to provide the first character of the Forward Sortation Area (FSA) of the postal codes of licensed premises but declined to provide the second and third characters or the names of the cities in which licensed production took place. At issue was whether these location data constituted “personal information” – which the government cannot disclose under s. 19(1) of the Access to Information Act (ATIA). A second issue was the degree of effort required of a government department or agency to maximize the release of information in a privacy-protective way. Essentially, this case is about “the appropriate analytical approach to measuring privacy risks in relation to the release of information from structured datasets that contain personal information” (at para 2).

The licensing scheme was available to those who wished to grow their own marijuana for medical purposes or to anyone seeking to be a “designated producer” for a person in need of medical marijuana. Part of the licence application required the disclosure of the medical condition that justified the use of medical marijuana. Where a personal supply of medical marijuana is grown at the user’s home, location information could easily be linked to that individual. Both parties agreed that the last three characters in a six-character postal code would make it too easy to identify individuals. The dispute concerned the first three characters – the FSA. The first character represents a postal district. For example, Ontario, Canada’s largest province, has five postal districts. The second character indicates whether an area within the district is urban or rural. The third character identifies either a “specific rural region, an entire medium-sized city, or a section of a major city” (at para 12). FSAs differ in size; StatCan data from 2016 indicated that populations in FSAs ranged from no inhabitants to over 130,000.

Information about medical marijuana and its production in a rapidly evolving public policy context is a subject in which there is a public interest. In fact, Health Canada proactively publishes some data on its own website regarding the production and use of medical marijuana. Yet, even where a government department or agency publishes data, members of the public can use the ATI system to request different or more specific data. This is what happened in this case.

In his decision, Justice Pentney emphasized that both access to information and the protection of privacy are fundamental rights. The right of access to government information, however, does not include a right to access the personal information of third parties. Personal information is defined in the ATIA as “information about an identifiable individual” (s. 3). This means that all that is required for information to be considered personal is that it can be used – alone or in combination with other information – to identify a specific individual. Justice Pentney reaffirmed that the test for personal information from Gordon v. Canada (Health) remains definitive. Information is personal information “where there is a serious possibility that an individual could be identified through the use of that information, alone or in combination with other available information.” (Gordon, at para 34, emphasis added). More recently, the Federal Court has defined a “serious possibility” as “a possibility that is greater than speculation or a ‘mere possibility', but does not need to reach the level of ‘more likely than not’” (Public Safety, at para 53).

Geographic information is strongly linked to reidentification. A street address is, in many cases, clearly personal information. However, city, town or even province of residence would only be personal information if it can be used in combination with other available data to link to a specific individual. In Gordon, the Federal Court upheld a decision to not release province of residence data for those who had suffered reported adverse drug reactions because these data could be combined with other available data (including obituary notices and even the observations of ‘nosy neighbors’) to identify specific individuals.

The Information Commissioner argued that to meet the ‘serious possibility’ test, Health Canada should be able to concretely demonstrate identifiability by connecting the dots between the data and specific individuals. Justice Pentney disagreed, noting that in the case before him, the expert opinion combined with evidence about other available data and the highly sensitive nature of the information at issue made proof of actual linkages unnecessary. However, he cautioned that “in future cases, the failure to engage in such an exercise might well tip the balance in favour of disclosure” (at para 133).

Justice Pentney also ruled that, because the proceeding before the Federal Court is a hearing de novo, he was not limited to considering the data that were available at the time of the ATIP request. A court can take into account data made available after the request and even after the decision of the Information Commissioner. This makes sense. The rapidly growing availability of new datasets as well as new tools for the analysis and dissemination of data demand a timelier assessment of identifiability. Nevertheless, any pending or possible future ATI requests would be irrelevant to assessing reidentification risk, since these would be hypothetical. Justice Pentney noted: “The fact that a more complete mosaic may be created by future releases is both true and irrelevant, because Health Canada has an ongoing obligation to assess the risks, and if at some future point it concludes that the accumulation of information released created a serious risk, it could refuse to disclose the information that tipped the balance” (at para 112).

The court ultimately agreed with Health Canada that disclosing anything beyond the first character of the FSA could lead to the identification of some individuals within the dataset, and thus would amount to personal information. Health Canada had identified three categories of other available data: data that it had proactively published on its own website; StatCan data about population counts and FSAs; and publicly available data that included data released in response to previous ATIP requests relating to medical marijuana. In this latter category the court noted that there had been a considerable number of prior requests that provided various categories of data, including “type of license, medical condition (with rare conditions removed), dosage, and the issue date of the licence” (at para 64). Other released data included the licensee’s “year of birth, dosage, sex, medical condition (rare conditions removed), and province (city removed)” (at para 64). Once released, these data are in the public domain, and can contribute to a “mosaic effect” which allows data to be combined in ways that might ultimately identify specific individuals. Health Canada had provided evidence of an interactive map of Canada published on the internet that showed the licensing of medical marijuana by FSA between 2001 and 2007. Justice Pentney noted that “[a]n Edmonton Journal article about the interactive map provided a link to a database that allowed users to search by medical condition, postal code, doctor’s speciality, daily dosage, and allowed storage of marijuana” (at para 66). He stated: “the existence of evidence demonstrating that connections among disparate pieces of relevant information have previously been made and that the results have been made available to the public is a relevant consideration in applying the serious possibility test” (at para 109). Justice Pentney observed that members of the public might already have knowledge (such as the age, gender or address) of persons they know who consume marijuana that they might combine with other released data to learn about the person’s underlying medical condition. Further, he notes that “the pattern of requests and the existence of the interactive map show a certain motivation to glean more information about the administration of the licensing regime” (at para 144).

Health Canada had commissioned Dr Khaled El Emam to produce and expert report. Dr. El Emam determined that “there are a number of FSAs that are high risk if either three or two characters of the FSA are released, there are no high-risk FSAs if only the first character is released” (at para 80). Relying on this evidence, Justice Pentney concluded that “releasing more than the first character of an FSA creates a significantly greater risk of reidentification” (at para 157). This risk would meet the “serious possibility” threshold, and therefore the information amounts to “personal information” and cannot be disclosed under the legislation.

The Information Commissioner raised issues about the quality of other available data, suggesting that incomplete and outdated datasets would be less likely to create reidentification risk. For example, since cannabis laws had changed, there are now many more people cultivating marijuana for personal use. This would make it harder to connect the knowledge that a particular person was cultivating marijuana with other data that might lead to the disclosure of a medical condition. Justice Pentney was unconvinced since the quantities of marijuana required for ongoing medical use might exceed the general personal use amounts, and thus would still require a licence, creating continuity in the medical cannabis licensing data before and after the legalization of cannabis. He noted: “The key point is not that the data is statistically comparable for the purposes of scientific or social science research. Rather, the question is whether there is a significant possibility that this data can be combined to identify particular individuals.” (at para 118) Justice Pentney therefore distinguishes between the issue of data quality from a data science perspective and data quality from the perspective of someone seeking to identify specific individuals. He stated: “the fact that the datasets may not be exactly comparable might be a problem for a statistician or social scientist, but it is not an impediment to a motivated user seeking to identify a person who was licensed for personal production or a designated producer under the medical marijuana licensing regime” (at para 119).

Justice Pentney emphasized the relationship between sensitivity of information and reidentification risk, noting that “the type of personal information in question is a central concern for this type of analysis” (at para 107). This is because “the disclosure of some particularly sensitive types of personal information can be expected to have particularly devastating consequences” (at para 107). With highly sensitive information, it is important to reduce reidentification risk, which means limiting disclosure “as much as is feasible” (at para 108).

Justice Pentney also dealt with a further argument that Health Canada should not be able to apply the same risk assessment to all the FSA data; rather, it should assess reidentification risk based on the size of the area identified by the different FSA characters. The legislation allows for severance of information from disclosed records, and the journalists argued that Health Canada could have used severance to reduce the risk of reidentification while releasing more data where the risks were acceptably low. Health Canada responded that to do a more fine-grained analysis of the reidentification risk by FSA would impose an undue burden because of the complexity of the task. In its submissions as intervenor in the case, the Office of the Privacy Commissioner suggested that other techniques could be used to perturb the data so as to significantly lower the risk of reidentification. Such techniques are used, for example, where data are anonymized.

Justice Pentney noted that the effort required by a government department or agency was a matter of proportionality. Here, the data at issue were highly sensitive. The already-disclosed first character of the FSA provided general location information about the licences. Given these facts, “[t]he question is whether a further narrowing of the lens would bring significant benefits, given the effort that doing so would require” (at para 181). He concluded that it would not, noting the lack of in-house expertise at Health Canada to carry out such a complex task. Regarding the suggestion of the Privacy Commissioner that anonymization techniques should be applied, he found that while this is not precluded by the ATIA, it was a complex task that, on the facts before him, went beyond what the law requires in terms of severance.

This is an interesting and important decision. First, it reaffirms the test for ‘personal information’ in a more complex data society context than the earlier jurisprudence. Second, it makes clear that the sensitivity of the information at issue is a crucial factor that will influence an assessment not just of the reidentification risk, but of tolerance for the level of risk involved. This is entirely appropriate. Not only is personal health information highly sensitive, at the time these data were collected, licensing was an important means of gaining access to medical marijuana for people suffering from serious and ongoing medical issues. Their sharing of data with the government was driven by their need and vulnerability. Failure to robustly protect these data would enhance vulnerability. The decision also clarifies the evidentiary burden on government to demonstrate reidentification risk – something that will vary according to the sensitivity of the data. It highlights the dynamic and iterative nature of reidentification risk assessment as the risk will change as more data are made available.

Indirectly, the decision also casts light on the challenges of using the ATI system to access data and perhaps a need to overhaul that system to provide better access to high-quality public-sector information for research and other purposes. Although Health Canada has engaged in proactive disclosure (interestingly, such disclosures were a factor in assessing the ‘other available data’ that could lead to reidentification in this case), more should be done by governments (both federal and provincial) to support and ensure proactive disclosure that better meets the needs of data users while properly protecting privacy. Done properly, this would require an investment in capacity and infrastructure, as well as legislative reform.

Artificial intelligence (AI) is already being used to assist government decision-making, although we have little case law that explores issues of procedural fairness when it comes to automated decision systems. This is why a recent decision of the Federal Court is interesting. In Barre v. Canada (Citizenship and Immigration) two women sought judicial review of a decision of the Refugee Protection Division (RPD) which had stripped them of their refugee status. They raised procedural fairness issues regarding the possible reliance upon an AI tool – in this case facial recognition technology (FRT). The case allows us to consider some procedural fairness guideposts that may be useful where evidence derived from AI-enabled tools is advanced.

The Decision of the Refugee Protection Division

The applicants, Ms Barre and Ms Hosh, had been granted refugee status after advancing claims related to their fear of sectarian and gender-based violence in their native Somalia. The Minister of Public Safety and Emergency Preparedness (the Minister) later applied under s. 109 of the Immigration and Refugee Protection Act to have that decision vacated on the basis that it was “obtained as a result of directly or indirectly misrepresenting or withholding material facts relating to a relevant matter”.

The Minister had provided the RPD with photos that compared Ms Barre and Ms Hosh the applicants) with two Kenyan women who had been admitted to Canada on student visas shortly before Ms Barre and Ms Hosh filed their refugee claims (the claims were accepted in 2017). The applicants argued that the photo comparisons relied upon by the Minister had been made using Clearview AI’s facial recognition service built upon scraped images from social media and other public websites. The Minister objected to arguments and evidence about Clearview AI, maintaining that there was no proof that this service had been used. Clearview AI had ceased providing services in Canada on 6 July 2020, and the RPD accepted the Minister’s argument that it had not been used, finding that “[a]n App that is banned to operate in Canada would certainly not be used by a law enforcement agency such as the CBSA” (at para 7). The Minister had also argued that it did not have to disclose how it arrived at the photo comparisons because of s. 22 of the Privacy Act, and the RPD accepted this assertion.

The photo comparisons were given significant weight in the RPD’s decision to overturn the applicants’ refugee status. The RPD found that there were “great similarities” between the photos of the Kenyan students and the applicants, and concluded that they were the same persons. The RPD also considered notes in the Global Case Management System to the effect that the Kenyan students did not attend classes at the school where they were enrolled. In addition, the CBSA submitted affidavits indicating that there was no evidence that the applicants had entered Canada under their own names. The RPD concluded that the applicants were Kenyan citizens who had misrepresented their identity in the refugee proceedings. It found that these factual misrepresentations called into question the credibility of their allegations of persecution. It also found that, since they were Kenyan, they had not advanced claims against their country of nationality in the refugee proceedings, as required by law. The applicants sought judicial review of the decision to revoke their refugee status, arguing that it was unreasonable and breached their rights to procedural fairness.

Judicial Review

Justice Go of the Federal Court ruled that the decision was unreasonable for a number of reasons. A first error was allowing the introduction of the photo comparisons into evidence “without requiring the Minister to disclose the methodology used in procuring the evidence” (at para 31). The Minister had invoked s. 22 of the Privacy Act, but Justice Go noted that there were many flaws with the Minister’s reliance on s. 22. Section 22 is an exception to an individual’s right of access to their personal information. Justice Go noted that the applicants were not seeking access to their personal information; rather, they were making a procedural fairness argument about the photo comparisons relied upon by the Minister and sought information about how the comparisons had been made. Section 22(2), which was specifically relied upon by the Minister, allows a request for disclosure of personal information to be refused on the basis that it was “obtained or prepared by the Royal Canadian Mounted Police while performing policing services for a province or municipality…”, and this circumstance simply was not relevant.

Section 22(1)(b), which was not specifically argued by the Minister, allows for a refusal to disclose personal information where to do so “could reasonably be expected to be injurious to the enforcement of any law of Canada or a province or the conduct of lawful investigations…” Justice Go noted that case law establishes that a court will not support such a refusal on the basis that because there is an investigation, harm from disclosure can be presumed. Instead, the head of an institution must demonstrate a “nexus between the requested disclosure and a reasonable expectation of probable harm” (at para 35, citing Canadian Association of Elizabeth Fry Societies v. Canada). Exceptions to access rights must be given a narrow interpretation, and the burden of demonstrating that a refusal to disclose is justifiable lies with the head of the government institution. Justice Go also noted that “the Privacy Act does not operate “so as to limit access to information to which an individual might be entitled as a result of other legal rules or principles”” (at para 42) such as, in this case, the principles of procedural fairness.

Justice Go found that the RPD erred by not clarifying what ‘personal information’ the Minister sought to protect; and by not assessing the basis for the Minister’s 22 arguments. She also noted that the RPD had accepted the Minister’s bald assertions that the CBSA did not rely on Clearview AI. Even if the company had ceased offering its services in Canada by July 6, 2020, there was no evidence regarding the date on which the photo comparisons had been made. Justice Go noted that the RPD failed to consider submissions by the applicants regarding findings by the privacy commissioners of Canada, BC, Alberta and Quebec regarding Clearview AI and its activities, as well as on the “danger of relying on facial recognition software” (at para 46).

The Minister argued that even if its s. 22 arguments were misguided, it could still rely upon evidentiary privileges to protect the details of its investigation. Justice Go noted that this was irrelevant in assessing the reasonableness of the RPD’s decision, since such arguments had not been made before or considered by the RPD. She also observed that when parties seek to exempt information from disclosure in a hearing, they are often required at least to provide it to the decision-maker to assess. In this case the RPD did not ask for or assess information on how the investigation had been conducted before deciding that information about it should not be disclosed. She noted that: “The RPD’s swift acceptance of the Minister’s exemption request, in the absence of a cogent explanation for why the information is protected from disclosure, appears to be a departure from its general practice” (at para 55).

Justice Go also observed that information about how the photo comparisons were made could well have been relevant to the issues to be determined by the RPD. If the comparisons were generated through use of FRT – whether it was using Clearview AI or the services of another company – “it may call into question the reliability of the Kenyan students’ photos as representing the Applicants, two women of colour who are more likely to be misidentified by facial recognition software than their white cohorts as noted by the studies submitted by the Applicants” (at para 56). No matter how the comparisons were made – whether by a person or by FRT technology – some evidence should have been provided to explain the technique. Justice Go found it unreasonable for the RPD to conclude that the evidence was reliable simply based upon the Minister’s assertions.

Justice Go also found that the RPD’s conclusion that the applicants were, in fact, the two Kenyan women, was unreasonable. Among other things, she found that the decision “failed to provide adequate reasons for the RPD’s conclusion that the two Applicants and the two Kenyan students were the same persons based on the photo comparisons” (at para 69). She noted that although the RPD referenced ‘great similarities’ between the women in the two sets of photographs, there were also some marked dissimilarities which were not addressed. There simply was no adequate explanation as to how the conclusion was reached that the applicants were the Kenyan students.

The decision of the RPD was quashed and remitted to be reconsidered by a differently constituted panel of the RPD.

Ultimately, Justice Go sends a clear message that the Minister cannot simply advance photo comparison evidence without providing an explanation for how that evidence was derived. At the very least, then, there is an obligation to indicate whether an AI technology was used in the decision-making process. Even if there is some legal basis for shielding the details of the Minister’s methods of investigation, there may still need to be some disclosure to the decision-maker regarding the methods used. Justice Go’s decision is also a rebuke of the RPD which accepted the Minister’s evidence on faith and asked no questions about its methodology or probity. In her decision, Justice Go take serious note of concerns about accuracy and bias in the use of FRT, particularly with racialized individuals, and it is clear that these concerns heighten the need for transparency. The decision is important for setting some basic standards to meet when it comes to reviewing evidence that may have been derived using AI. It is also a sobering reminder that those checks and balances failed at first instance – and in a high stakes context.