On 12 November 2026 the New York City Council passed Introduction 891, which prohibits pricing set "completely or in part by algorithm that uses personal data to offer different prices to different customers for the same goods or services," according to the Council's legislative record. The vote was 44-5. The Department of Consumer and Worker Protection (DCWP) enforces the ban, with civil penalties of up to $1,000 per violation. The Record reports that Seattle signed a similar ban last month, making New York the second city to pass one. The state's own bill is still unsigned on Gov. Kathy Hochul's desk.
The strongest case for the ban
The case for the ban is stronger than its critics usually admit. Ordinary price discrimination, such as student discounts or coupons, is visible to the buyer. Personalised pricing built on a person's data is not. A shopper who sees $12 has no way to know that someone else saw $9, or that the quote reflected their device, location or browsing history. Markets discipline prices only when buyers can compare, and individualised quotes remove that comparison.
The Federal Trade Commission documented the infrastructure behind this. Its January 2025 staff findings found that intermediary firms worked with "at least 250 clients" selling everything from groceries to apparel. The Commission voted 3-2 to release the findings, and two commissioners dissented. The study was still ongoing, and the staff summaries relied on hypothetical examples because 6(b) data is published only in anonymised form. The FTC's findings show that the tooling exists and is widely sold. They do not show how often it raises prices for individual shoppers. That gap matters for what follows.
What the bill gets right
The bill is narrower than the label "surveillance pricing" suggests. It does not cap prices or ban dynamic pricing as such. The Council's hearing announcement says the legislation excludes loyalty and rewards programs, publicly disclosed discounts, and pricing differences tied to the cost of providing a good or service. It also says the bills do not prohibit discounts or promotions, naming senior and military discounts and coupons. The Record adds buy-one-get-one deals and student and veteran discounts to the list of carve-outs.
Those exemptions keep the bill aimed at the conduct that actually erodes trust: charging a specific person more because of what is known about them. A $1,000 per-violation penalty is meaningful for a small merchant but unlikely, on its own, to change the behaviour of a large platform. Each affected transaction or customer can count as a separate violation, so exposure can scale for a large seller.
A ban also fits better than a disclosure rule. New York's existing disclosure law, General Business Law § 349-a, requires a notice reading "THIS PRICE WAS SET BY AN ALGORITHM USING YOUR PERSONAL DATA." As the Norton Rose Fulbright analysis records, a federal court dismissed the National Retail Federation's First Amendment challenge on 8 October 2025, and the law took effect on 10 November 2025. A notice tells shoppers the practice exists but does not stop it, and few people can act on it at the checkout. The Council's move to prohibition suggests lawmakers saw disclosure as an incomplete remedy.
Where the risk lies
The weak point is the definition. "Pricing set in part by an algorithm that uses personal data" is broad. Personal data is not only a browsing history. It can include a delivery address, a payment method or a loyalty identifier. A delivery fee that varies by address, or a price that varies by store location, reflects real cost differences. Yet it could be argued to fall inside the bill's wording. The exemption for pricing tied to the cost of providing a good or service is meant to protect these cases. The Council's page does not say how that line will be drawn in practice.
The legislative record also describes a second exemption, for "certain regulated activity," without saying what it covers. Businesses cannot plan around an exemption they cannot read. DCWP should publish guidance and examples before enforcement begins, and the Council should expect to revisit the text once the first complaints arrive. Proportionate regulation means penalising the targeting of individuals, not every pricing model that touches a data field.
There is also an evidence problem. The ban rests on the premise that firms are setting individual prices from personal data to the buyer's detriment, but the public record so far is mostly about capability. If DCWP gathers data on actual price differences and publishes it, the next round of rules can rest on evidence rather than inference.
The patchwork question
Consumer Reports analyst Grace Gedye told The Record that the bill would spare eight million New Yorkers from guessing whether their search history shapes the price they see. She also said businesses that must change their practices for New York City may extend those changes nationwide. That outcome would be convenient, but it is not guaranteed. The Record notes that New Jersey, Connecticut and Maryland have similar bans, and that the New York legislature passed its own bill in June. A city rule, a state rule and Seattle's rule that differ in definition and enforcement will push compliance costs onto smaller sellers that cannot afford a legal team.
The cleaner path is alignment. Hochul should sign a state law with a definition that matches the city's, so a retailer faces one standard in New York. Both measures should keep their exemptions for transparent promotions. A federal standard would be better still, but that is unlikely in the near term.
Bottom line
Banning individualised pricing from personal data is a defensible rule, and it is narrower than its critics claim. Its success will depend on whether DCWP applies the definition to targeting and not to ordinary cost-based pricing, and on whether New York's state and city rules end up consistent.