The rise of artificial intelligence allows businesses to maximize profitability through analyzing data to personalize prices–changing individual prices according to consumers’ willingness to pay. As advancements in data collection and analysis continue to progress, a more formal model of personalized pricing, in which each consumer receives their own individual price, is becoming increasingly viable.¹ Would this have a positive effect on overall welfare or be skewed towards firms? The following essay argues that personalized pricing facilitated through new technology risks pushing markets towards first-degree price discrimination–where firms reduce consumer surplus through charging consumers close to the maximum price they are willing to pay. Widespread price discrimination risks engendering negative consumer perceptions of fairness, which can harm consumer satisfaction and trust. However, there are potential upsides: in the right transparent and competitive market, this pricing model can benefit the overall allocative efficiency of the economy while also allowing consumers access to goods at lower prices, benefiting consumer welfare. This essay first examines potential consumer harms, then allocative benefits, before demonstrating how the true outcome is determined by the market and regulatory environment.

The use of artificial intelligence in the analysis of various forms of data (past purchases, location, search history, etc.) allows businesses to make predictive claims about consumers’ willingness to spend. Companies such as Delta Air Lines have already sparked debates around personalized prices with the integration of artificial intelligence into their variable pricing models, in contrast to traditional pricing models that set standard fixed percentage mark-ups to the cost of an item or service.² With the computing power for artificial intelligence growing, algorithms gaining sophistication, and increasing pressures for firms to integrate AI, personalized pricing in the digital age has reached a level of efficacy that warrants a closer examination of the potential risks and benefits.³

As prices increase and consumers pay closer to the maximum price they are willing to pay, consumer surplus decreases. Consequently, the highly effective personalized pricing strategies that have been adopted can create a significant reduction in consumer surplus.

At its core, consumer surplus is a measure of consumer benefit. But it also holds significant implications for overall consumer welfare and business strategies. If surplus is large, then consumers enjoy greater savings, meaning that, in general, they are more willing to make purchases that maintain a high consumer surplus; that incentivizes businesses to work to develop products and pricing strategies to create more perceived value.⁴ However, the correlation between consumer surplus and net welfare is not necessarily a given as necessities like food and healthcare break the equation given that they are relatively inelastic–possessing similar demand at every price. For necessities, willingness to pay, represented in a high consumer surplus, can reflect desperation or unavoidability, not real welfare.

Relating consumer surplus to the debate around personalized pricing, when businesses price discriminate–charging different prices based on individual preferences and factors–customer perceptions of fairness can be negatively affected. Specifically concerning new developments in artificial intelligence, studies show that consumers perceive the use of AI or data algorithms in determining personalized prices negatively. The rationale for a consumer’s negative reaction is that being charged a higher price than someone else for the same product is inequitable and unfair, which causes consumers to lose trust in a company.⁵

In the past, companies have already experimented with personalized prices such as in a 2012 Wall Street Journal investigation that found corporations like Staples and Home Depot raising prices based on personal factors like geolocation, income level, and proximity to rival stores. In all, these examples show how companies can and will use demographic and geographical data to extract maximal profits from consumers behind the scenes.⁶ Furthermore, cases of potential price discrimination involving companies like Amazon also confirm that consumers reacted negatively when discovering that prices were individually inflated. In this instance, Amazon quickly refunded and denied the use of personalized pricing, claiming it to be a random price test.⁷ The backlash shows that the connection between consumer satisfaction and personalized pricing is a concern that firms must take real steps to alleviate or face consumer dissatisfaction, which can negatively affect costs and profitability.

The important takeaway remains that consumer satisfaction is fragile and highly susceptible to being affected by personalized pricing. While consumer harms are certainly a significant issue to focus on, policymakers inevitably have to deal with trade-offs, necessitating a look at potential upsides to personalized pricing.

One of the biggest economic upsides to personalized pricing practices becoming widespread would be an increase in allocative efficiency. Similar to the variable pricing model of personalized pricing, dynamic pricing is a separate, distinct example of a model currently being used to regulate demand that can provide insight into the benefits of personalized pricing. Companies such as Uber utilize surge pricing, the elevation of prices as demand suddenly increases, an example of how non-static pricing models like personalized pricing can be utilized as a tool for allocative efficiency. For instance, customers who urgently need a ride will be more likely to pay the inflated price, whereas customers who can wait will not be, thereby using pricing as a tool to allocate the limited resource to the consumer who needs it the most.⁸

Similarly, personalized pricing more closely aligns the cost of a good or service with an individual consumer’s willingness to pay, regulating demand. The benefit further extends to consumers through an increase in competition driven by firms collectively utilizing personalized pricing to undercut their competitors, which creates downward pressure on prices.⁹ For the same reason, personalized pricing can increase consumer surplus in cases that a customer couldn’t afford to pay for a product at a certain price as companies can recognize this and will lower prices to maximize sales.¹⁰

A model created by Toulouse Professor of Economics Andrew Rhodes and Yale Professor of Economics Jidong Zhou expands on these findings, demonstrating that when market coverage is sufficiently high–when production costs are low or there are many competing firms–then personalized pricing reduces net profits and increases consumer surplus, suggesting a net benefit to consumer welfare.¹¹ This can be seen empirically through examples like airlines offering targeted discounts such as student rates to customers with a lower willingness to pay, who otherwise might not have been able to afford tickets.¹²

However, these positive effects are incredibly dependent on the competition environment that personalized pricing models exist in; if the market is highly competitive, then another competitor will step in to undercut high prices, preventing inequitable pricing structures that charge customers unfair amounts.¹³

In the event of a non-competitive market, two distinct potential problems arise: traditional abuse of dominance and selective pricing. For the first case, a dominant firm could use personalized pricing as a tool for exploitative abuse, making prices excessive or unfair to consumers who lack recourse due to the superior bargaining position of the monopolistic firm. The second case presents when a dominant firm abuses personalized pricing as an exclusionary tool, meant to target rivals’ consumers by offering them lower prices to cut their competitors out of the market.¹⁴

For example, a firm could steal a rival’s customer by offering a product for which a customer has a lesser preference at a much lower personalized price with the hope of decreasing their rival’s market share and potentially putting them out of business. This effect can happen when some firms have access to more or better customer information and thus are able to apply personalized pricing to greater effect. In that instance, the same model by Rhodes and Zhou shows that total welfare and match efficiency decrease as a single firm poaches customers who aren’t their best fit.¹⁵

Both cases prevent the benefits of a more affordable personalized price by either having the opposite effect, raising prices, or creating the conditions to unfairly raise prices by eliminating competition that would regulate prices from getting too high. This ties back into the exception to consumer surplus and welfare as this effect is particularly harmful for inelastic goods provision, as consumers are functionally forced into paying unaffordable and extreme prices. Overall, the variability in success of personalized pricing models suggests that the net effect is determined more by the market structure and regulatory environment surrounding it.

Nevertheless, given the above-mentioned benefits, there is a place for some personalized pricing: as a tool to increase competition, allocative efficiency, and consumer access to goods at lower price points. But to avert the aforementioned risks, best practices should be followed.

First, businesses must be held to a standard of transparency, disclosing relevant information about pricing strategies to the public. This requirement should be coupled with comprehensive enforcement and investigation to prevent attempts to capitalize off personalized pricing in unfair and misleading ways.¹⁶ Both measures address consumer satisfaction concerns as described above and attempt to solve for any disparities between different firms’ ability to enact personalized pricing by making any proprietary data collection techniques public. In this last case, policies that require more open data sharing would allow for personalized pricing while removing the issue of exclusion highlighted above.¹⁷

Secondly, rigorous enforcement and amendment of antitrust laws surrounding personalized pricing must be instituted, applying extra scrutiny to firms that maintain significant market influence.¹⁸ This is important to prevent collusion–a practice when competing producers work together to undermine competitive processes such as price-setting. Collusion presents a unique problem for personalized pricing as the stakes are elevated given the ability to raise all market prices to the theoretical maximum a consumer would pay without competitive recourse in a market with firms colluding.¹⁹

Finally, data privacy laws and the regulation of the collection and use of consumer data would have to be more stringent. With a focus on consent, new regulations would doubly improve transparency by directly disclosing the data collected and giving individuals the opportunity to amend and change aspects of their personalized profile, among other stringent limits to how data is utilized.²⁰ This recommendation to regulate data privacy works synergistically to the first recommendation by dispelling potential privacy concerns of companies openly sharing consumer data by controlling the data companies can possess in the first place.

Personalized pricing presents itself as a unique and pressing issue for policymakers: it is not inherently harmful, as it can be a tool to improve allocative efficiency and expand access to goods given proper guidelines and stringently enforced regulations. However, it is important to realize the risks to consumer satisfaction, surplus, and competitive markets that widespread personalized pricing poses. The emphasis on market structure and regulatory environment for net outcome exposes just how contingent the effectiveness of the recommendations is on political will. This ultimately poses a greater question for the framework for addressing the emergence of personalized pricing: should addressing long-term issues like consumer welfare prevail over short-term maximization of profit?

Notes

1. Miroslava Marinova and Christian Bergqvist, “AI-Enabled Price Discrimination as an Exploitative Abuse of Dominance under EU Competition Law,” Journal of Competition Law & Economics (2026): nhag006, https://doi.org/10.1093/joclec/nhag006. ↩
2. Jay L. Zagorsky, “Personalized pricing has spread across many industries. Here's how consumers can avoid it,” PBS News, August 3, 2025, https://www.pbs.org/newshour/economy/personalized-pricing-has-spread-across-manyindustries-heres-how-consumers-can-avoid-it. ↩
3. Benjamin Shiller, “Buyer beware: Does AI-powered personalized pricing actually help consumers? Brandeis economist weighs in,” interview by Julian Cardillo, Brandeis Stories, August 12, 2025, https://www.brandeis.edu/stories/2025/august/shiller-ai-pricing.html. ↩
4. Dara-Abasi Ita, “Consumer Surplus: Definition, Measurement, and Example,” Investopedia, May 20, 2025, https://www.investopedia.com/terms/c/consumer_surplus.asp. ↩
5. Xiao Peng, Xixian Peng, and David (Jingjun) Xu, “Does AI Disclosure in Discriminatory Pricing Backfire? The Moderating Role of Price Sensitivity and Explanation for Price Differences,” Proceedings of the 57th Hawaii International Conference on System Sciences (2024): 6858, https://scholarspace.manoa.hawaii.edu/server/api/core/bitstreams/e14a9c25-b05f-4a75-985213d0500d8ea2/content. ↩
6. OECD Secretariat, “Personalised Pricing in the Digital Era,” Directorate for Financial and Enterprise Affairs Competition Committee (November 2018): 16, https://www.oecd.org/content/dam/oecd/en/publications/reports/2018/10/personalised-pricing-in-the-digitalera_7313c12d/db4d9c9c-en.pdf. ↩
7. OECD Secretariat, “Personalised Pricing,” 16. ↩
8. Christopher Gardner and Juan Londoño, “Personalized Pricing Isn’t All Bad for Consumers,” Cato Institute, May 28, 2026, accessed 6/18/26, https://www.cato.org/blog/personalized-pricing-isnt-all-bad-consumers. ↩
9. Shiller, “Does AI-powered personalized pricing actually help consumers?” ↩
10. Gardner and Londoño, “Personalized Pricing Isn’t All Bad.” ↩
11. Andrew Rhodes and Jidong Zhou, “Personalized Pricing and Competition,” American Economic Review vol. 114, no. 7 (July 2024): 2141-70, https://www.aeaweb.org/articles?id=10.1257/aer.20221524. ↩
12. Arian Aflaki and Qian (Kenneth) Zhang, “Personalizing the Prices While Keeping Customers Happy,” The FinReg Blog, January 9, 2023, https://sites.duke.edu/thefinregblog/2023/01/09/personalizing-the-prices-while-keepingcustomers-happy/. ↩
13. Gardner and Londoño, “Personalized Pricing Isn’t All Bad.” ↩
14. OECD Secretariat, “Personalised Pricing,” 28. ↩
15. Andrew Rhodes and Jidong Zhou, “Personalized Pricing and Competition,” 25. ↩
16. OECD Secretariat, “Personalised Pricing,” 7. ↩
17. Rhodes and Zhou, “Personalized Pricing and Competition,” 26. ↩
18. OECD Secretariat, “Personalised Pricing,” 7. ↩
19. Gardner and Londoño, “Personalized Pricing Isn’t All Bad.” ↩
20. OECD Secretariat, “Personalised Pricing,” 39. ↩