Four states have enacted laws banning the practice of charging individual consumers different prices based on their personal data — and on Tuesday, Congress held its first dedicated hearing on the subject, with senators from both parties calling it one of the worst consumer scams in American history. But the most alarming testimony at the August 4 Senate Judiciary subcommittee session wasn't about what surveillance pricing algorithms are already doing to shoppers. It was about what comes next: AI agents with real-time access to your emotional state, your desperation, and the urgency of your purchase — and no federal law on the horizon that specifically covers them.

AI Has Already Reinvented the Ripoff

Surveillance pricing is not hypothetical. Target's app raises prices when it detects a customer is physically inside the store. DoorDash has developed technology that adjusts recommendations based on how hungry it estimates a user to be. Travel sites like Expedia have charged users from high-cost cities more for identical hotel rooms. And Kroger's loyalty program uses customer shopping data to infer household income.

A January 2025 FTC surveillance pricing study confirmed that at least 250 retail clients were using third-party pricing intermediaries — companies hired specifically to track consumer behavior and set individualized prices. Those intermediaries catalog mouse movements on webpages, the contents of abandoned shopping carts, real-time location data, and demographic inferences. Their outputs: a per-customer price, calculated to extract the maximum that person will pay before walking away.

At Tuesday's hearing in Dirksen Senate Office Building Room 226, Lindsay Owens, president and CEO of Groundwork Collaborative, called the result a fundamental reinvention of consumer exploitation. "Companies can now purchase, track, and analyze your personal data at a previously unimaginable scale and clip," Owens told the subcommittee. "They catalog your location, your purchase history, and even your cursor movements to better predict exactly how much they can charge you. Even nominally beneficial programs, like loyalty rewards, can be used to harvest your information, turning customers into guinea pigs in sophisticated pricing experiments."

The documented cost is already real. A December 2025 Consumer Reports Instacart pricing investigation found that Instacart's AI pricing experiments caused identical items to cost up to 23 percent more for some shoppers than for others buying from the same store at the same time. For a typical family of four, the researchers estimated the practice adds roughly $1,200 to annual grocery bills.

Bipartisan Wrath, a Familiar Problem

Republicans and Democrats don't agree on much. Tuesday's hearing was a notable exception.

Sen. Josh Hawley (R-MO), who chairs the Subcommittee on Crime and Counterterrorism, opened by targeting Amazon, Walmart, Kroger, Lyft, Staples, and Target. He cited reports that Kroger made roughly $500 million by tracking its customers in-store and online and selling that data to third parties — and noted one case in which a Kroger customer received a 62-page copy of the shopping profile the company had assembled about him. Kroger, in a statement provided separately, denied the premise: "Kroger does not use personal data, actual or inferred, to personalize or raise prices."

Hawley presented data on Lyft that illustrated the gap between company denials and documented practice: 55 separate riders in Missouri were charged 29 different prices for the same route at the same time.

"This is one of the biggest scams in American history — this marriage between the AI industry and these mega corporations," Hawley told the subcommittee. "And they're trying to pull it off against every American consumer. AI surveillance pricing is the unholy trinity of everything Americans hate: spying on people, ripping them off, and taking away jobs."

Sen. Dick Durbin (D-IL), the subcommittee's ranking Democrat, was, by his own characterization, in full agreement. He said he had little to add after Hawley's opening, and the bipartisan session pressed expert witnesses on what surveillance pricing is and what Congress should do about it.

Robert Hedges, a digital fellow at MIT and former chief data officer at Visa, provided the session's most technical testimony: a walkthrough of the specific data flows that make surveillance pricing possible, including payment-network transaction data. "Willingness to pay — that's the expression," Hedges told the subcommittee, describing how the combination of location data, ZIP codes, IP addresses, purchase histories, and behavioral dwell time on product pages is used to identify each individual's reservation price. The fact that credit card transaction data — aggregated across every merchant where a consumer shops — flows into pricing systems was the detail that most clearly illustrated the scope of the surveillance architecture.

The one dissenting voice came from John Zhang, a professor of marketing at the Wharton School of Business at the University of Pennsylvania. Zhang argued that collecting consumer data is necessary to understand the market and that personalized pricing, properly applied, can benefit consumers through targeted discounts. His position drew polite but firm pushback from multiple senators.

The Agentic AI Escalation No Legislation Currently Addresses

The most consequential part of Tuesday's hearing was a scenario Owens sketched that described not today's surveillance pricing problem but tomorrow's.

Imagine a parent of a sick child at midnight, asking Walmart's AI shopping assistant, Sparky, to suggest a thermometer and some Tylenol. Sparky knows the purchase is urgent. It has access to the user's purchase history, their conversations, the time of day, and the emotional context of the request. A standard surveillance pricing algorithm profiles past behavior to predict willingness to pay. An agentic AI commerce system — one in an active conversation with you — can do something categorically more dangerous: it can detect real-time desperation and exploit it at the moment of sale.

"As agentic commerce reshapes how we shop online, with AI agents gaining access to your habits, purchases, and personal conversations, the risk of being ripped off will only grow," Owens told the hearing.

This is the structural gap that makes Tuesday's hearing more urgent than a simple policy update. Every bill currently in committee — the One Fair Price Act introduced by Sen. Ruben Gallego (D-AZ), the bipartisan No Rigged Grocery Prices Act from Reps. Josh Gottheimer (D-NJ) and Mike Lawler (R-NY), and the forthcoming Hawley legislation he said after the hearing he expects to introduce "soon" — targets data collection and algorithm design. None specifically regulates what an AI agent is permitted to infer from the emotional content and timing of a consumer's real-time request. Today's legislation is built to address yesterday's surveillance pricing architecture.

Owens' conclusion was direct: "The answer can't be to teach everyone how to beat the machine. We must set rules of the road so the machine plays fair."

How the Price Is Set Against You

Surveillance pricing algorithms operate on a concept called the "reservation price" — the maximum a specific buyer will pay before abandoning a purchase. The goal of first-degree price discrimination (the economic term for what surveillance pricing attempts to achieve) is to charge every consumer exactly their reservation price, capturing all value that would otherwise remain with the buyer as consumer surplus.

The data inputs that feed reservation-price models include:

• Real-time location (inside a store vs. comparison shopping at home signals different price sensitivity)
• Device type (iPhone users have historically been shown higher prices than Android users)
• Time of day and urgency signals (a midnight purchase for a child's medicine reads as inelastic demand)
• Purchase history and browsing patterns (abandoned carts signal near-purchase; return customers signal loyalty and therefore willingness to pay more)
• Inferred income (derived from neighborhood data, payment method, and credit card transaction history)
• Payment network data (Visa and Mastercard transaction histories provide cross-merchant spending profiles)

Robert Hedges' testimony on the credit card data input was the detail most senators found alarming: the surveillance pricing ecosystem does not depend solely on the data a single retailer collects. It draws on data aggregated across the entire commercial landscape, sold and re-sold by brokers and pricing intermediaries.

"Uncovering surveillance pricing requires comparing prices across users, devices, and locations all at once," Owens told the hearing. "This is a hard task for veteran researchers and a nearly impossible one for ordinary consumers."

Four State Bans, One Federal Patchwork Problem

While Congress deliberated Tuesday, four states had already moved from deliberation to law.

Maryland signed the nation's first food-sector surveillance pricing ban on April 28, 2026. The law applies to food retailers over 15,000 square feet and third-party delivery platforms; it will take effect October 1, 2026, with fines of $10,000 to $25,000 per violation.

Connecticut enacted a broader ban on May 27, 2026, covering all retail sellers. Its distinctive provision: when a price has been raised based on personal data, retailers must display the label "THIS PRICE WAS INCREASED BY A PRICE SETTING DEVICE USING YOUR PERSONAL DATA." Connecticut's ban also takes effect October 1, 2026.

New York's legislature passed the One Fair Price Act on June 4, 2026, sending it to Governor Kathy Hochul for her signature. The bill, championed by Attorney General Letitia James, would replace New York's existing disclosure-only regime (in effect since November 2025) with an outright prohibition. As of publication, Hochul had not yet signed the bill into law.

New Jersey became the fourth state when Gov. Mikie Sherrill signed the Fair Price Protection Act on July 23, 2026. New Jersey's law is structurally different from the others: it gives individual consumers the right to sue retailers themselves, without having to petition the attorney general. For willful violations, courts may award treble damages — three times actual harm — which makes the law viable even for individually small overcharges when pursued as a class action. The NJ law takes effect February 1, 2027.

Colorado's legislature passed a broader ban in May 2026 that would have covered all industries and included a private right of action, but Gov. Jared Polis vetoed it on June 2, arguing the bill was overly broad and could inadvertently restrict legitimate consumer discounts.

The patchwork creates compliance complexity for retailers with multistate operations — and an obvious gap for the roughly 300 million Americans who live outside the four states with enacted protections.

Federal Enforcement Is Moving on Two Tracks

The Federal Trade Commission is not waiting for legislation. The agency issued an April 2026 advance rulemaking notice on April 14, covering personalized pricing disclosure, fee transparency, and unauthorized billing protections for online food and grocery delivery. A coalition of 16 state attorneys general — led by New York and Tennessee — sent the FTC a letter in May 2026 urging it to both extend its existing fee rule and issue a separate rule targeting surveillance pricing.

The Robinson-Patman Act, the 1936 federal law barring price discrimination in sales between competing buyers, has also returned to active enforcement after decades of dormancy. The FTC brought its first Robinson-Patman Act suits in a generation in late 2024 and early 2025, and the Department of Justice signaled in June 2026 that criminal enforcement of algorithmic pricing collusion — the practice of multiple retailers using shared pricing-algorithm vendors to achieve tacit price coordination without explicit agreement — remains on the table.

The DOJ settled with RealPage in November 2025, restricting data sharing among clients of a shared algorithmic pricing vendor in the rental housing market — a precedent that applies directly to the multi-retailer pricing intermediary model now common in grocery and retail.

Hawley said after Tuesday's hearing that he expects to introduce his own surveillance pricing legislation soon. The session marked the first time the Senate has formally treated AI-powered personalized pricing as a dedicated hearing subject, rather than folding it into broader inquiries on pricing, AI, or data privacy.

Will Congress Act?

Whether the bipartisan energy in Room 226 translates into legislation is the central question. Congress has a poor recent track record on technology regulation when powerful commercial interests are arrayed in opposition — and the retail, grocery, delivery, and airline industries have substantial lobbying capacity.

Amazon, Walmart, Kroger, JetBlue, Lyft, and Target did not appear as witnesses at Tuesday's hearing. Walmart and Kroger, through spokesperson statements issued separately, denied using personal data to raise individual prices — claims senators and expert witnesses placed in direct tension with the documented pricing experiments, patents, and data-brokerage practices the hearing examined.

The companies' absence left the legislative record largely uncontested. Owens' framing — that Congress must set rules of the road so the machine plays fair — provided the hearing's closing logic: not whether algorithmic pricing exists, but whether it will be governed by rules built for the age of agentic AI, or by rules built for an era of static behavioral profiling that has already passed.

Frequently Asked Questions

What is AI surveillance pricing, and how is it different from a regular sale price?

A store sale lowers the price for everyone who walks in. Surveillance pricing calculates the highest price a specific individual is likely to pay — based on their location, browsing history, payment data, inferred income, and now their real-time emotional state — and charges them that amount instead of the market price. The economic concept is first-degree price discrimination: extracting the maximum value from each transaction individually. AI makes it scalable; agentic AI makes it situationally aware.

What do current state laws actually protect me from, and when do they take effect?

As of August 2026, three states have enacted signed bans: Maryland and Connecticut bans take effect October 1, 2026, covering food retailers and all retail sellers respectively. New Jersey's ban takes effect February 1, 2027, and is the only state law that gives you a personal right to sue a retailer — without waiting for the attorney general — if you believe you were overcharged based on your data. New York passed its One Fair Price Act through both legislative chambers on June 4, 2026, but Governor Hochul had not yet signed it into law as of publication. If you live outside the three states with enacted protections, no specific prohibition applies at the federal level.

Why do the existing bills not cover agentic AI shopping assistants?

Current federal proposals — including the One Fair Price Act and the No Rigged Grocery Prices Act — target data collection and algorithmic pricing design: they restrict what personal data companies may use to set prices. Agentic AI commerce, however, can exploit the content and urgency of a real-time request — "I need a thermometer for my sick child right now" — in a way that is distinct from profiling past behavior. The AI agent doesn't need your historical data to know you're desperate in this moment. Existing bills do not specifically regulate what inferences an AI agent may draw from an active conversation, and no proposed legislation closes that gap as of August 2026.

What can I do right now to limit surveillance pricing exposure?

Use a private browsing window or clear cookies before significant purchases; pricing systems often rely on tracked session data. Compare prices across different devices (desktop vs. mobile) and browsers — the same item can show different prices depending on device. A VPN masks your location, one of the most heavily weighted pricing inputs. For grocery delivery apps, compare prices between the app and the store's own website before ordering. In Maryland and Connecticut (starting October 1, 2026), New York (in effect), and New Jersey (from February 2027), contact your state attorney general if you believe you've been charged more than other customers for the same item based on your personal data.