New Jersey just became the third US state to ban surveillance pricing. The headlines make it sound like AI pricing is under attack. It isn’t, but only if you can explain the difference. Here’s what pricing executives need to tell their boards this week.
A headline that scares the wrong people
On July 23, New Jersey Governor Mikie Sherrill signed the Fair Price Protection Act, making New Jersey the third US state to ban what regulators call “surveillance pricing.” Maryland and Connecticut got there first. New York’s version is sitting on the governor’s desk. And more than 40 algorithmic-pricing bills have been introduced across at least 24 states this year.
If you run pricing for a retailer with US exposure, your inbox probably filled up with some version of the same question: are we still allowed to do this?
The honest answer is that most retailers are asking about the wrong thing. The laws moving through these legislatures are narrow and specific. They do not ban dynamic pricing. They do not ban AI in pricing. They ban one practice, and confusing that practice with legitimate price optimization is how good teams talk themselves into unnecessary panic, or worse, into quietly switching off tools they were fully entitled to use.
What the New Jersey law actually prohibits
Read past the headline, and the New Jersey law does one thing. It prohibits businesses from using a shopper’s personal data, their browsing activity, location, device, and purchase history, to charge that individual a different price than the next person for the identical product.
That is surveillance pricing. It is a specific, identity-based practice: two people, same item, same moment, different prices, because the retailer knows something about each of them. It is what regulators, labor unions, and consumer advocates have spent two years worried about, and it is what these laws are built to stop.
The New Jersey statute is also careful about what it leaves alone. It does not touch loyalty programs. It does not touch standard discounts. And it includes a separate, one-year pause on the rollout of new electronic shelf labels while the state studies their impact; existing labels can stay in place, be repaired, and be replaced.
None of that describes how modern dynamic pricing works.
The distinction every pricing leader needs to be able to say out loud
There are two very different things that both get loosely called “algorithmic pricing,” and the entire regulatory conversation turns on telling them apart.
Personalized pricing sets a price for a person. It takes individual consumer data as an input and produces a price aimed at that specific shopper. This is what the New Jersey, Maryland, and Connecticut laws prohibit.
Dynamic pricing sets a price for a product. It responds to market conditions, competitor moves, demand, cost changes, inventory levels, elasticity, and shows the resulting price to everyone looking at that product at that moment. There is no individual shopper in the equation. The price on the shelf is the price on the shelf, for you and for the person standing next to you.
A retailer using dynamic pricing to keep pace with competitors and protect margin is not doing what these laws prohibit. A pricing executive who cannot articulate that difference to their legal team, their board, or a state attorney general is exposed, not because their pricing is illegal, but because they cannot demonstrate that it isn’t.
Where the real risk sits
The legal risk in 2026 is not “we use algorithms.” It is “we cannot explain what our algorithms do.”
That is the throughline connecting New Jersey to the enforcement actions and inquiry letters emerging across the country. Regulators are asking a consistent set of questions: What data feeds your pricing? Can you explain why a specific product got a specific price on a specific day? Is that decision documented? Does competitor data enter your system in a way that could look like coordination?
A pricing tool built as a black box cannot answer those questions. It can tell you the price it recommended. It cannot reliably tell you why. And the exposure does not stop with the pricing software, the retailer deploying the tool carries the compliance risk, whatever the contract says.
So the retailers most at risk from this regulatory wave are not the ones doing dynamic pricing. They are the ones doing any kind of algorithmic pricing without an explainability layer, an audit trail, or documented human governance over the decisions.
What to do before this reaches your state
The practical steps are the same whether or not your state has passed a law yet, because the direction of travel is unmistakable.
Separate personalization from optimization in your own stack. Confirm, in writing, that your pricing engine does not use individual consumer data to set individualized prices. If it does, that is a live compliance problem in three states and counting.
Make your pricing explainable. For any given product and date, you should be able to produce a human-readable account of why a price was recommended and what inputs drove it. If the answer is “sort of,” that gap is where the legal exposure lives.
Check how your pricing software uses competitor data. Pooled competitor data (one shared price dataset that steers multiple competing retailers) used as a direct pricing input is exactly what several of these laws are designed to scrutinize. Understand what your tool relies on before a regulator asks.
Document your governance. Approval workflows, override controls, and escalation paths show that accountable humans oversee your pricing. That matters as compliance posture and as evidence of good faith.
How Quicklizard thinks about this
Quicklizard is an AI-powered dynamic pricing platform built on what we call Glass Box AI, the principle that every pricing recommendation needs to be explainable, auditable, and traceable to the logic and data that produced it.
That principle lines up directly with where the law is heading. Quicklizard optimizes on market signals, competitor prices, demand, cost, inventory, not on the identity or personal data of individual shoppers. The price our system recommends is the same for every customer looking at that product. That is dynamic pricing, and it sits outside the scope of the surveillance-pricing bans now moving through US statehouses.
Just as important, our customers can see and explain every recommendation: what inputs drove it, what rules governed it, and which person signed off. Pricing workflows include human approval steps, override capability, and documented decision authority. And Quicklizard does not use pooled competitor data in ways that create the anticompetitive-coordination exposure several of the new laws specifically target. Customers have seen up to a 15% revenue lift and a 11% profit increase, delivered through pricing logic they can see, explain, and stand behind.
The New Jersey law is not a threat to retailers who price responsibly and transparently. It is a reason to make sure you are one of them.
If you want to pressure-test your pricing infrastructure against the regulations taking effect this autumn, speak with a Quicklizard pricing strategist about your readiness.
