What: AI-driven dynamic pricing is being marketed to food retailers with claims of 2 to 3 percentage points of margin gain, while Canada's three grocery majors publicly rule out the practice.

Why it is important:  The margin case for AI pricing is now quantified and vendor-ready, which shifts the decision from technical feasibility to a deliberate choice between short-term yield and price-trust positioning. 

The pitch came from a Quebec firm at the ALL IN conference in Montreal, which claimed to have lifted a global retailer's profit margins by 2 to 3% through AI-assisted forecasting and dynamic pricing, replacing a strategy that ignored local sensitivities and short-term competitive moves with hourly per-store forecasting. Vendor staff framed the application defensively, citing regional markdowns on slow-selling trousers or celery nearing waste, both standard inventory management. The concern is what happens when the same system identifies a profitable pattern, such as raising rotisserie chicken prices on busy evenings in young-family neighbourhoods or lifting raspberry prices when a competitor runs out of stock.

Metro, Loblaw and Sobeys all confirmed that no dynamic pricing is in place online or in store, and that they do not intend to introduce it, with Loblaw citing customer expectations of fair and consistent prices. Walmart did not respond. In the United States, a Consumer Reports investigation found Instacart prices varying by up to 23% between customers, a practice since discontinued, and a bill now targets electronic shelf labels and surveillance pricing. Canada's NDP motion against personalised dynamic pricing was rejected the same day it was tabled.

IADS Notes: The Quebec vendor's pitch sits inside a debate already well documented. Electronic shelf labels have been defended as pure accuracy infrastructure, incapable of surge pricing or customer tracking, according to NRF in March 2026, yet the same hardware underpins Walmart's machine-learning pricing patents and its rollout across 4,600 US stores, reported by the Financial Times in March 2026 at a moment when several state legislatures were proposing grocery dynamic-pricing bans. Where prices are set from personal data rather than from local demand, the Financial Times noted in May 2026 that the practice attracts both consumer backlash and regulatory attention, and New York has since banned it outright, as Reuters covered in June 2026. The commercial conclusion drawn by The Robin Report in June 2026 is that the reputational and legal exposure now outweighs the margin gain, which is precisely the trade-off the Canadian grocers have resolved in public.

Could AI push up prices right before you buy dinner?