Surveillance pricing, or the use of personal information to set different prices for different consumers, is often treated as a novel problem requiring new legislation. The debate is further muddied by noise about how regulation could inadvertently harm consumers, including by restricting legitimate discounts.

However, Canada already has a well-established precedent for regulating against unfair discriminatory pricing in the Telecommunications Act of 1993. We have simply failed to apply that framework to surveillance pricing.

Subsection 27(1) of the Act requires that all Canadian carriers of telecommunication services must charge rates that are just and reasonable. Subsection 27(2) prohibits carriers from charging a rate that is unjustly discriminatory or subjects any person to an undue or unreasonable disadvantage.

This principle is also not unique to the Telecommunications Act, but in fact has long been a staple of transport, energy, and communication regulation. For example, the Broadcasting Distribution Regulations under the Broadcasting Act of 1991 contains a very similar provision on undue preference or disadvantage.

Surveillance pricing can be understood in similar terms, too. It is effectively a form of discrimination that may subject certain consumers or segments to unreasonable disadvantages by charging them prices that are arbitrarily, and often unknowably, higher than those paid by others.

Canada already has a framework to prohibit discriminatory practices

Following policy progress in the US, Canada is beginning to engage in the broader policy conversation around data-driven pricing strategies that segment consumers based on increasingly sophisticated profiles. Yet despite a commitment in the federal AI strategy to modernize privacy laws and to “ensure that Canadians’ personal information is not used inappropriately, including for surveillance pricing,” the recently tabled privacy bill, Bill C-36, does not explicitly address the practice. The bill provides more clarity on what may count as an inappropriate purpose and imposes stronger transparency requirements around automated decision-making, but it does not specify how these provisions apply to surveillance pricing.

So Canada still seems to be waffling around on how to respond to these practices—or even if we should deal with them at all. But Canadian law already articulates a relevant guiding principle. The principle of just and reasonable pricing should not be confined to telecommunications; it needs to be applied more broadly across the economy.

As important, those provisions of the Telecommunications Act have been tested. Subsection 27(2) grants the Canadian Radio and Telecommunications Commission (CRTC) broad and highly discretionary authority to address unjust discrimination, a power it has exercised repeatedly.

For example, in 2015, the CRTC found Bell (one of the country’s largest telecommunications providers) had violated subsection 27(2) for effectively self-preferencing its own products through its pricing practices. The complaint alleged that as a result of Bell self-preferencing its own services, it effectively charged customers eight times more to access competitors’ content offerings.

Surveillance pricing is a different phenomenon, and the basis for discrimination is different. Yet the effect is the same: consumers are penalized by a firm that can leverage its position to engage in unjust discrimination and charge them more than others. In the case of surveillance pricing, that leverage derives from the excessive amounts of data that firms collect and use to profile consumers.

Some have expressed concerns that regulating surveillance pricing will eliminate discounts based on broad demographic categories, such as student or senior discounts. But a prohibition on unfair price discrimination need not have that effect. There is an essential difference between universal discounts applied transparently based on a single voluntarily disclosed characteristic, like a person’s age, and individualized pricing based on a consumer’s inferred willingness to pay, determined from an obtuse array of personal data points.

Quebec offers us another example of legislative precedent for addressing surveillance pricing. Law-25 prohibits the use of personal information for profiling unless explicitly authorized and also applies to de-identified data, where federal privacy laws remain comparatively weak. Ostensibly, requirements related to profiling would apply to surveillance pricing, but they have yet to be tested in court. At the very least, the law reflects an established principle: some uses of personal information are inappropriate, regardless of their commercial value.

These two relatively modest provisions, one in long-established federal legislation and the other in a provincial privacy law, offer a clear framework for addressing surveillance pricing. Taken together, they suggested Canada is further ahead on this issue than it may appear. Yet that progress is not reflected in the newly proposed federal privacy legislation.

Why privacy law leaves the issue unresolved

Surveillance pricing is enabled by the collection and exploitation of consumers' personal information. At its core, it is a privacy issue. Accordingly, it is best tackled through privacy legislation.

Under the current federal framework, subsection 5(3) of the Personal Information Protection and Electronic Documents Act (PIPEDA) permits organizations to collect, use, or disclose personal information for purposes that are appropriate. Guidance from the Privacy Commissioner sets out that the use of personal information for profiling or categorization that leads to unfair, unethical, or discriminatory treatment contrary to human rights law is not allowed.

However, the bar the law sets is quite high—only discrimination based on grounds prohibited in human rights law would be considered inappropriate. The Commissioner’s guidance also suggests that unfair or unethical profiling will “generally be found inappropriate,” though there is no definition of what this might mean in practice.

This is where applying the principle in the Telecommunications Act against unjust discrimination could enable PIPEDA to explicitly prohibit the use of personal information for unfair discrimination. By amending the law to define unfair discrimination to include pricing strategies that use personal information to unjustly set individualized prices for different consumers.

Bill C-36 does expand on what constitutes an appropriate purpose and provides a list of factors organizations must take into account, including whether an individual’s loss of privacy is proportionate to the benefits of using the personal data. Fairness and non-discrimination are not among those factors. Whether surveillance pricing would be considered an inappropriate purpose is unclear, though additional obligations could be prescribed in regulations made under the act.

Another strength of the Telecommunications Act’s approach is that it provides the regulator with broad discretion to determine whether obligations have been met, recognizing that there will be important variations on a case-by-case basis. Likewise, an amended PIPEDA could provide the Privacy Commissioner with similar discretion for determining when the use of personal information constitutes unfair discrimination, and ideally, the teeth it needs to enforce against infractions.

Transparency is not enough 

What Canada ought not rely on, however, is transparency-based regulation of surveillance pricing. As various US states have moved to regulate surveillance pricing, many have looked to New York’s Algorithmic Pricing Disclosure Act as a model. The law requires that firms using surveillance pricing prominently display a warning that discloses that a price was set using an algorithm based on a consumer’s personal information.

Bill C-36 proposes a weaker disclosure requirement: organizations must make information available about their uses of any automated decision-making system that could have a significant impact on an individual. The bill does not require that such a warning be prominently disclosed rather than buried in a dense privacy policy, and it leaves the threshold for what counts as a “significant effect” undefined.

Transparency has value. It is important for consumers to know when and how their information is being used. But disclosure alone will not prevent firms from engaging in surveillance pricing. The fundamental issue is not a lack of transparency; it is the largely unchecked collection and exploitation of the public’s data. The fact that it mostly occurs without a consumer’s knowledge only compounds its effect, but it is not the source of it.

As the Telecommunications Act exemplifies, if we were unwilling to accept disclosure as a remedy for unfair discrimination in pricing access to telecommunications services —why should we accept the practice of widespread surveillance pricing, where the impacts will be felt even more broadly?

Still, starting with how personal information is used is essential. Firms have always looked for novel ways to exploit consumers, and limits on invasive data collection provide an important check on their ability to further do so, especially as AI is enabling companies to utilize and leverage data at an unprecedented scale.

Canada has established a strong precedent for fair and non-discriminatory pricing through legislation and regulations. While surveillance pricing may be a privacy problem, the debate should fundamentally turn on values. We must take the clear and principled stance that using personal information to set individualized prices is unfair discrimination and regulate it accordingly.