When you walk through the international departures terminal at Toronto Pearson on any busy Saturday morning, you’ll see an environment that hasn’t significantly changed in its shopping logic for several decades. You’ll see the families negotiating luggage carts, the quiet shuffle of passengers through duty-free with the unique combination of purpose and aimlessness that airport shopping tends to produce, and check-in lines that stretch back. You look through the items on the shelf. You choose what appeals to you.
At a register, you make your payment. Other than the fact that your passport indicates your origin and your flight departs in three hours, the airport knows very nothing about you. It is anticipated that by the end of 2026, the entire experience will look very different. AI systems are driving this change because they have a far better understanding of what a particular traveler is likely to want before they have even passed the first display case.
| Category | Details |
|---|---|
| Topic | AI-Driven Tax-Free Shopping in Canadian Airports |
| Target Completion | End of 2026 (full implementation expected) |
| Lead Airport | Toronto Pearson (major digital renovation underway) |
| AI Shopping Model | “Agentic Commerce” — AI agents facilitate automated purchases |
| Key Features | Hyper-personalization, virtual assistants, click-and-collect, dynamic pricing |
| Revenue Driver | Non-aeronautical revenue — retail optimization per passenger |
| Technology Tools | AR mirrors, 3D product visualization, multilingual voice AI |
| Operational Benefit | Predictive stock management, real-time inventory tracking |
| Investment Trend | Private investor interest in high-end airport retail space |
| Consumer Shift | Personalized, frictionless shopping becoming baseline expectation |
| Reference Website | catsa-acsta.gc.ca |
In order to create a shopping environment that functions more like a finely tuned digital storefront than a traditional duty-free floor, Canadian airports are moving toward what the industry refers to as AI-driven tax-free retail. This model combines passenger data, such as flight details, nationality, travel history, and past purchasing patterns, with real-time inventory management and personalized recommendation systems.
The commercial reasoning is fairly straightforward: airports receive a large amount of their non-aeronautical revenue from retail, and per-passenger spending rates that are significant at scale can be used to measure the difference between a traveler who wanders duty-free and buys nothing and one who receives a specific, personalized offer for something they were actually considering purchasing before they arrived.
The more ambitious end of what’s being implemented is the “agentic commerce” approach, which holds that AI agents facilitate purchasing decisions and execute transactions on behalf of travelers rather of just providing possibilities. When a mobile app is linked to a traveler’s trip and profile, it may instantly process payment, allow pre-ordering before the passenger arrives at the terminal, and send a notification suggesting a product with a particular discount.
The purchase is waiting to be picked up. On the way to the gate, the traveler picks it up. Previously the sole model for airport retail, physical browsing is now elective rather than required, which alters who purchases and what they purchase in ways that the data is already starting to measure. The infrastructure investment that makes this type of real-time customisation operationally feasible is Toronto Pearson’s digital transformation, which moves away from the old systems that have limited how passenger data can be used and integrated.
When explained to passengers, the dynamic pricing dimension elicits the most conflicting emotions. The efficiency logic of AI systems that modify prices in real time depending on stock levels, demand trends, or even the duration of a flight delay is difficult to contest from the retailer’s point of view, but the implications for the customer experience are more nuanced.
A traveler will react in a particular way if they discover that the perfume they were considering has increased in price by 10% due to a two-hour flight delay to Tokyo and an unusually busy airport. Depending on the traveler’s position in the transaction, the response may be “the system is working as designed” or “the system is taking advantage of a captive audience.” Therefore, airports that use dynamic pricing must carefully consider how they explain what the system is doing and why.
An additional element that contributes to the explanation of why this shift is occurring so rapidly is the private investment factor. For airport operators looking to diversify away from reliance on airline connections, non-aeronautical revenue—the money they make from retail, food and beverage, parking, and commercial activities rather than landing fees and gate charges—has emerged as the main growth sector.
The AI-driven customization strategy is a viable way to close the gap that private investors see in Canadian airport retail space, which has been structurally failing its digital potential for years. Because the investment thesis is clear—better data leads to higher per-passenger spending, and higher per-passenger spending leads to better returns on the retail space investment—capital is accessible to finance digital remodeling and AI retail infrastructure.
There’s a sense that the window for retailers that haven’t started integrating AI into their operations is smaller than they may have thought when looking at the extent of what’s planned for Canadian airport shopping by the end of 2026. When travelers encounter a truly personalized airport shopping environment at one terminal, their tolerance for the standard browse-and-hope experience at other terminals will decrease correspondingly. Passengers’ expectations for personalized, frictionless shopping are being set by experiences in other contexts, such as e-commerce, streaming, and food delivery.