Artificial intelligence could soon enable powerful companies to charge different customers different prices for the same product, based on predictions about what each individual is willing to pay. That is the warning from new research co-authored by Miroslava Marinova at the University of East London. The study argues that the central concern is not only the potential for higher prices, but the rise of hidden, personalised pricing that consumers cannot easily detect or understand.

Traditionally, firms have set prices in response to broad market forces such as demand, production costs, and competition. In this model, consumers are typically offered similar prices for the same product at any given time, creating a sense of transparency and predictability. While discounts or promotions may vary, the baseline expectation has long been that pricing is broadly consistent across customers.

A different model is now taking shape. Algorithmic personalised pricing uses data-driven systems to tailor prices at the level of the individual consumer. Rather than responding only to overall market demand, these systems aim to estimate how likely a specific person is to accept a higher price instead of searching for alternatives. In effect, pricing decisions become personalised predictions about behaviour rather than general responses to market conditions.

AI systems can draw on vast amounts of data, including browsing history, location, purchasing patterns, and even device usage, to infer willingness to pay. As a result, the same product could be offered at different prices to different individuals at the same time. While forms of price differentiation have existed for years, artificial intelligence significantly increases the precision and scalability of this approach, bringing markets closer to a scenario in which every consumer is quoted a unique price.

The study, co-authored with Christian Bergqvist of the University of Copenhagen, highlights that the most pressing issue may not be price levels themselves but perceptions of fairness. Even if average prices remain stable, consumers tend to react strongly when they discover they are paying more than others without a clear or justified reason. This perceived inequity can erode trust and influence purchasing behaviour in significant ways.

In competitive markets, consumers may still have the option to switch to cheaper alternatives. However, the researchers note that where a dominant firm is involved, personalised pricing could raise legal concerns. In such cases, it may be interpreted as an abuse of market power under existing EU and UK competition law, particularly if the pricing lacks transparency or objective justification. The absence of visibility makes it difficult for consumers to assess whether they are being treated fairly.

The paper concludes that while legal frameworks already contain tools to address these issues, regulation has not yet fully caught up with the capabilities of AI-driven pricing. As these technologies become more widespread, policymakers and regulators will face growing pressure to determine where to draw the line. In the UK, where competition rules closely mirror those of the EU, there are already indications that authorities may expand oversight, including potential new powers for the Competition and Markets Authority to examine algorithmic pricing practices more closely.

More information: Miroslava Marinova et al, AI-Enabled Price Discrimination as an Exploitative Abuse of Dominance under EU Competition Law, Journal of Competition Law & Economics. DOI: 10.1093/joclec/nhag006

Journal information: Journal of Competition Law & Economics Provided by University of East London