As online shopping becomes increasingly embedded in everyday life, companies are gaining greater ability to tailor prices to individual consumers. Using personal information—including purchasing patterns, geographic location, demographic characteristics, online behaviour and even biometric data—businesses can determine what particular customers may be willing to pay. Artificial intelligence could dramatically accelerate this practice of surveillance-based pricing, further deepening the divide between what consumers believe remains private and how their personal information is actually being used.
Consider a parent who turns to an AI chatbot for advice on buying a thermometer. The chatbot responds helpfully, recommending a product and directing the user to a retailer. What the parent does not know is that information from earlier conversations about their sick child’s symptoms has been passed along to the seller. The retailer now knows that the customer is worried about a child with a rapidly rising fever—and therefore has an urgent need for the product. Using that vulnerability as a pricing signal, it raises the price of the thermometer.
If companies can combine chatbot conversations, browsing history, location, purchasing behaviour, demographics and biometric information, they may no longer need to offer everyone the same price. Instead, they could estimate each individual’s urgency, financial capacity and willingness to pay—and adjust prices accordingly. A consumer who urgently needs medicine, a thermometer, a flight home or another essential product could therefore face a higher price precisely because the system knows that delaying the purchase is difficult.
This fundamentally changes the nature of price discrimination. Traditional pricing strategies generally divide consumers into broad categories. AI makes it possible to move toward individual-level pricing, where the relevant market is effectively a single person and the price reflects what an algorithm believes that particular person can be persuaded or compelled to pay.
The deeper concern is therefore not simply that companies possess large quantities of personal data. It is that information collected in one context—such as asking an AI assistant about a sick child—could potentially be transformed into economic intelligence in another. Consumers may believe they are seeking information privately while, in practice, their circumstances could become signals about their willingness to pay.
That creates a serious imbalance between businesses and consumers. Sellers equipped with sophisticated AI systems may know far more about customers than customers know about the mechanisms determining the prices placed before them. Two people could consequently encounter different prices for the same product without knowing that the difference exists or understanding why.
The policy question, then, is whether personal data should be permitted to determine individualized prices at all—and, if so, what information may legitimately be used. Without meaningful transparency and privacy protections, AI could transform surveillance from a tool for targeted advertising into something more consequential: a mechanism for determining the price attached to each individual consumer.
