Algorithms know what you're willing to pay in an emergency

Imagine you're sitting at home, suffering through a heatwave in an unprepared flat. You would probably reach for your phone and look for fans or air conditioning (AC) devices, right? That's exactly what many did during July’s heatwave in France. Online retailers know this and act accordingly. When most of France was melting under temperatures that reached 40ºC, the average online price of fans increased by 46%, and AC by 22%. That's how capitalism works: when demand grows, prices go up. Consumers are therefore already used to seeing prices change quickly, but what is new is algorithms that draw on personal information to charge each customer as much as they're individually willing to pay. This is called “algorithmic pricing” and includes different practices, such as dynamic and personalised pricing, Frederik Zuiderveen Borgesius, professor of ICT and Law at Radboud University, explained to The European Correspondent. “Dynamic pricing is when the prices of products go up and down, sometimes rather quickly, according to demand or other factors, but [everyone] will still see the same price,” he told us. Personalised pricing is different. Consumers can be shown different values for the same product based on their data, Zuiderveen Borgesius said. “On rare occasions, you can observe personalised pricing if you and a friend visit an online store at the same time and see different prices.” Using our digital footprint, systems can try to estimate our 'willingness to pay' and “then try to match that to earn the maximum possible amount from a single person,” Joanna Strycharz, assistant professor at the University of Amsterdam, explained. The problem with personalised pricing is that it is harder to detect, according to a European Parliament study on personalised pricing from November 2022. For starters, online prices are often opaque. Unlike traditional taxis with stable and transparent tariffs or meters, consumers on ride-hailing apps simply receive a price that can change at any moment and without explanation. “What is the starting price? Why does it go higher or lower? What is the default? How much per kilometre?”, illustrated Strycharz. Uber has long faced questions over personalised pricing, but has denied using “protected characteristics, phone battery levels, phone models, or other device information to set prices.” Cost ceilings for emergencies Price personalisation can also be used in consumers' favor. For instance, “it may allow certain groups to purchase products that they could otherwise not afford” by offering them better prices, according to a report by the European Parliament. Unfortunately, there are examples where algorithmic pricing has been used against consumers in emergencies, known as price gouging. According to SER Catalunya, during the Iberian blackout in 2025, tariffs increased by 300% on Cabify, 116% on Uber and 30% on Bolt. Strycharz pointed to other controversies surrounding services, like during the 2014 Sydney hostage crisis, when Uber fares rose to four times normal rates, with a minimum fare of 100 Australian dollars, or during winter storm Nemo in Boston, where Uber fares doubled. According to Strycharz, this then led to proposals regarding certain goods or services that should not be allowed to increase their pricing above a certain amount during emergencies. After the floods in Valencia in 2024, the Spanish government changed the law so that prices could not be raised during situations of “emergency, risk or necessity”, but the amendment was derogated by Parliament last February due to lack of political support. Borgesius compares it with the prohibition of the abuse of circumstances. “If someone is drowning and you demand €1,000 before helping them, their desperate agreement does not necessarily make the deal legitimate.” According to him, an extremely high fare during an emergency could raise similar legal questions. Hide from algorithms So, is there a way to protect yourself against algorithmic pricing? Consumers have developed tricks that allegedly shield you from these practices: searching in incognito mode or deleting cookies to avoid being tracked, or leaving items in your basket to prompt retailers to send you discount codes. According to the experts, though, it is extremely tough to know whether those tricks actually work. “It’s often very difficult to spot personalised prices,” Zuiderveen Borgesius said. “Stores can personalise prices by sending some people discounts by email. In this way, price personalisation can be hard to observe. Hence, determining whether a price changed only for a particular consumer is often pretty much impossible,” he concluded. In 2025, the Dutch Authority for Consumers and Markets forced the online retailer Wish to stop using personalised pricing and fake discounts. What about EU laws, then? “EU consumer protection rules do not regulate the price levels for goods and services,” a European Commission spokesperson told us. “They also do not prohibit price changes, including when these are implemented via dynamic or personalised pricing. Traders can freely determine the prices they charge for their products.” There are, however, rules on how prices are set and what information consumers must be provided with. According to the Commission, traders must disclose when a price has been determined through personalisation. “Misleading and unfair commercial practices that deceive consumers, such as changing prices during the booking process or omitting important information, are also prohibited,” the spokesperson said. But for Strycharz, disclosure alone isn't enough: “Even if it says this price is personalised, you still wouldn’t know why and how.” Transparency, she argued, needs to go further, including requiring platforms to disclose what actually shapes a price or offer. This debate is entering the forthcoming Digital Fairness Act. Euroconsumers, an umbrella group of six European consumer associations, argues that existing rules do not adequately protect consumers against dynamic and personalised pricing. It's calling for consumers to be told explicitly when prices are dynamic, for prices to stop changing once a purchase has started, and for advertised prices to stay valid for a set period during an online session.