Companies have always tried to predict what customers will pay – airlines have long adjusted fares based on demand, and retailers have experimented with personalized prices. But AI is giving companies a much more powerful way to answer an old question:
Maxime Cohen, a professor at McGill University and author of the forthcoming MIT Press book Pricing in the Age of AI, has spent years studying how technology is changing what companies charge and how those decisions are made.
Cohen says AI is taking pricing from broad segmentation to much more granular personalization.
“Traditional dynamic pricing typically adjusts prices based on factors such as demand, inventory, or time, whereas AI can combine a large number of signals about an individual customer, including purchase history, browsing behaviour, location, and other contextual information, to estimate willingness to pay in real time. The result is the possibility of moving closer to the “right price for the right customer at the right time”,” he says.
Personalized pricing to grow
Cohen says he expects personalized pricing to grow significantly, but much of it may initially take the form of personalized discounts, promotions, and offers rather than different posted prices for every customer.
“Adoption will be fastest among large e-commerce retailers, marketplaces, travel companies, and loyalty-driven businesses that have rich customer data, frequent transactions, and the technological infrastructure to experiment at scale. The book emphasizes that sophisticated AI-based pricing is already particularly feasible for businesses with large customer bases, substantial data, and strong technological capabilities,” he says.
Charging two customers different prices is not inherently problematic; we already accept this in settings such as airlines, hotels, student discounts, and loyalty programs, explains Cohen.
“The concern arises when the difference feels unfair, exploitative, or discriminatory, particularly when consumers do not understand why they are paying more or when pricing relies on sensitive attributes or proxies for them. With AI, the key challenge is not simply whether prices differ, but whether those differences can be justified and explained as fair and transparent,” he says.
Big opportunity for retailers
The opportunity is enormous for retailers, says Cohen.
AI can help retailers respond faster to market conditions, better predict demand, target discounts more effectively, and enhance both revenue and profitability. But pricing is very sensitive because customers immediately notice when they feel they have been treated unfairly; an algorithm that produces a short-term profit increase can in some cases destroy long-term trust and loyalty if customers believe they are being exploited. The best pricing algorithm is not necessarily the one that extracts the highest price today, but the one that creates sustainable value over time,” he says.
Cohen says Generative AI could democratize sophisticated pricing because retailers will increasingly be able to ask an AI system to analyze products, markets, competitors, and customer contexts and recommend a price without the need of building a complex pricing algorithm from scratch.
“Looking further ahead, we may see the rise of agent-to-agent commerce, where consumers will delegate shopping decisions to AI agents that interact (and potentially negotiate) with retailers’ AI agents. As pricing becomes more autonomous on both sides of the transaction, the big question will shift from “Can AI recommend a price?” to “How much authority should we give AI to actually set and negotiate prices?”,” he says.
More from Retail Insider:
