Personalized Pricing: The Privacy Problem

Have you ever looked up flights in incognito mode, or on a new browser? Called a hotel directly to see what the real rates are? Compared ride-sharing prices on your friend’s phone?

“Algorithmic pricing” refers to the use of software and mathematical models to set or recommend dynamic prices, often in real time, based on a variety of data inputs. Functionally, algorithmic pricing systems continuously analyze a range of data sources, including consumer demographics, online behavior, transaction history, and broader market data such as supply, demand, and competitor pricing.

These algorithms draw inferences from the collected data to estimate each consumer’s likely willingness to pay and set or recommend optimal prices that may be adjusted dynamically as new information becomes available, which could include data—intentionally or not—that are associated with protected characteristics.

While there are obvious benefits for businesses, including the lack of direct human intervention, there are also a number of issues that can arise—antitrust concerns when algorithms collude, price-gouging for concert tickets, bizarre feedback loops driven by mindless machines, not to mention opacity and fairness concerns. When these data inputs are driven by personal data, algorithmic pricing becomes personal pricing, which implicates privacy, competition, and consumer protection law.

Old Competition Concerns, New Tools

Concerns about algorithmic pricing echo long‑standing competition law issues such as price‑fixing and tacit collusion. Canada’s Federal Court has described price-fixing as akin to fraud and theft, observing that the practices “represent nothing less than an assault on our open market economy.”

While algorithmic pricing avoids explicit agreements between competitors, similar risks arise as firms may rely on the same data sources or third‑party pricing services, enabling coordinated pricing behavior without overt collusion. Canada’s Competition Bureau has studied the issue, publishing a discussion paper and the results of their consultation, but has so far chosen to monitor the situation rather than take action against individualized, opaque, and data-driven pricing outcomes.

Public Skepticism and the Fairness Gap

Recent polling found that a majority of Canadians support regulating algorithmic pricing, with 52% in support of a full ban of the practice, and 31% in favour of stricter regulation. This skepticism appears closely linked to concerns about how personal data is collected and used, and it reflects a broader unease about the power imbalance between companies with access to vast amounts of data and individual consumers navigating increasingly complex digital marketplaces.

Consumers may be more accepting of price changes driven by observable market conditions than of personalized or discriminatory price based on opaque data-driven inferences. Demand‑based pricing feels less intrusive than individualized price differentiation informed by a “black box” of data inputs that are invisible to the consumer.

Algorithmic Pricing as a Stress Test for Privacy and PIPEDA

Federal privacy laws, alongside certain proposed provincial regulatory frameworks, may provide consumers with relief. As the federal government has signalled renewed efforts to modernize the Personal Information Protection and Electronic Documents Act (PIPEDA) in concert with its forthcoming AI strategy, algorithmic pricing offers a concrete illustration of how contemporary data practices can produce individualized economic effects that privacy frameworks currently struggle to address.

Algorithmic pricing exemplifies the kinds of AI‑driven harms that modernized privacy legislation must be equipped to confront as data uses proliferate. These practices highlight the need for safeguards governing automated decision‑making systems with material individual impact, draw attention to the significance of inferences as distinct from directly collected data, and challenge privacy frameworks that continue to conceptualize harm primarily through the lens of data breaches. In the context of algorithmic pricing, the use of data may remain formally lawful while nevertheless eroding consumer agency, fairness, and trust.

Privacy law may be best positioned to respond to these concerns because it directly regulates the collection, use, and inference of personal information through enforceable limits on consent and purpose. Much of the public discomfort stems from the lack of transparency behind algorithmic pricing systems, rather than price variability alone. Concerns about surveillance, profiling, and loss of control over personal information closely mirror the foundational principles of PIPEDA, which include consent, identifying purposes for collection, and openness.

A modernized PIPEDA could address these challenges by explicitly engaging with automated decision making that has material consumer impacts, imposing obligations around explainability and contestability that extend beyond notice, and incorporating proportionality‑based assessments of data use, including reliance on third‑party or brokered data.

Provincial Pushback from Manitoba and Yukon

Manitoba introduced the first legislation in Canada to address the issue of algorithmic pricing in 2026. The Business Practices Amendment Act broadly defines “personalized algorithmic pricing” as applicable both online and in physical retail stores, where electronic shelf labelling systems may be used. The Act deems such practices to be unfair and carry fines of up to $300,000 for a first offence and up to $1,000,000 for a subsequent offence.

Rental housing illustrates the risks of algorithmic pricing in markets where competition is highly localized. The use of such tools by a few landlords could restrict price competition for renters, as highlighted by the Canadian Anti-Monopoly Project’s submission to the Competition Bureau’s consultation on the issue. The Yukon government has responded to this issue with direct legislative intervention by making it an offence to set rental rates using algorithms under the new Residential Tenancies Act.

Efficiency Without Exploitation

Together, recent legislative developments and public reaction underscore that algorithmic pricing has become less a question of innovation, and more an issue of data governance and trust. Arguments in favour of algorithmic pricing emphasize potential efficiencies, including improved demand matching, inventory management, and innovation, but these benefits do not require unbounded data extraction or opaque surveillance practices. The viability of algorithmic pricing going forward will depend on whether those efficiencies can be delivered within privacy frameworks that impose meaningful limits on how personal data is collected, inferred, and used to shape individual economic outcomes. The challenge for regulators and businesses will be to ensure that technological efficiency does not come at the expense of transparency, fairness, and consumer trust.