Businesses have long used personal information to decide which advertisement, coupon or product recommendation a shopper sees. Regulators are now examining a more consequential use: helping determine the price itself.
A US Senate Judiciary subcommittee examined the issue on August 4 during a hearing titled “Your Data, Their Profit: The Consumer Cost of AI Surveillance Pricing.”
The Federal Trade Commission uses the term “surveillance pricing” for practices in which detailed information about consumers is used to help target prices or offers. Other researchers and businesses use terms such as personalized pricing or data-driven pricing, and there is no single definition accepted by every participant in the debate.
AI can make it easier to process large amounts of customer data and estimate how price-sensitive someone may be. However, the practice does not always require advanced AI, and not every price calculated by an algorithm is personalized.
It is different from ordinary dynamic pricing
Prices have always changed for reasons such as demand, inventory, season, location and the cost of providing a service.
A hotel may charge more during a major event, while a ride-hailing company may raise prices when the number of passengers exceeds the number of available drivers. Those are forms of dynamic pricing.
The more controversial practice begins when information connected to a particular customer or device helps determine the price or promotion that person sees.
Two shoppers could therefore receive different offers for the same product, not simply because they visited at different times, but because a system placed them into different customer categories.
What information could affect the price?
In 2024, the Federal Trade Commission began studying companies that provide pricing technology and data services to retailers.
The FTC’s initial findings, published in January 2025, said these systems can use information including:
- A customer’s location and demographic profile
- Browsing and shopping history
- Previous purchases
- The device, time and sales channel used
- Mouse movements on a webpage
- Products left unpurchased in an online shopping cart
That information can be used to divide customers into groups and tailor prices, discounts or product selections to them.
However, the FTC study focused on pricing intermediaries rather than producing a public list of retailers using each practice. Because the agency had to protect confidential business information, its published examples were hypothetical.
The findings therefore show that the technology and data systems exist. They do not establish how often every capability is used, which retailers use it or how many customers have received higher prices because of their personal data.
Why businesses may find it attractive
A single price forces a company to balance two goals. A high price may produce a larger profit on each sale but discourage price-sensitive shoppers. A lower price may attract more customers but reduce the margin on people who would have paid more.
Personalized pricing attempts to manage both groups separately.
In testimony submitted for the Senate hearing, Wharton marketing professor Z. John Zhang argued that personalized pricing is not automatically harmful. He said it could allow companies to offer lower prices to more price-sensitive customers, increase sales and make some products available to people who might otherwise be priced out.
The same capability could also be used in the opposite direction: identifying customers who appear willing to pay more and offering them a higher price or a smaller discount.
The technology does not guarantee that either companies or consumers will benefit. The outcome depends on how the system is designed, the level of competition in the market and whether customers can compare prices elsewhere.
The risk is not limited to privacy
For businesses, personalized pricing creates a customer-trust problem as well as a data-privacy issue.
Shoppers are familiar with clearly explained discounts for loyalty-program members, students, older consumers or first-time buyers. The terms are usually visible, and people can understand why the price is different.
A hidden system based on browsing behavior, location or inferred willingness to pay is harder to explain. A customer who discovers that someone else received a better price may feel that the company used personal information against them.
This makes pricing an increasingly important part of brand management. A system may improve revenue in the short term while damaging trust if customers believe the process is secretive or unfair.
New rules are beginning to appear
New York already requires companies covered by its Algorithmic Pricing Disclosure Act to notify consumers when a price is set by an algorithm using their personal data.
The required notice states: “THIS PRICE WAS SET BY AN ALGORITHM USING YOUR PERSONAL DATA.”
The law took effect on November 10, 2025. It requires disclosure rather than imposing a general ban on all personalized pricing, and it includes several exceptions.
The legal picture is also becoming more complicated across the country. A Future of Privacy Forum review counted dozens of privacy and algorithmic pricing bills introduced during 2026 as of March. The proposals ranged from disclosure requirements to restrictions covering particular industries or uses of personal data.
Pricing is becoming a data-governance decision
Companies considering personalized pricing must look beyond whether an algorithm increases sales or profit.
They also need to know which customer information enters the system, whether that information is accurate, whether different groups are treated fairly and whether the resulting price can be explained to a consumer or regulator.
The important shift is that customer data is no longer used only to choose the message shown to a shopper. It may also help choose the price attached to the product.
For businesses, that turns pricing from a revenue-management decision into a question involving privacy, compliance, customer trust and brand reputation.
