Algorithmic pricing disclosure law, read from the statutes.

Two states require a sentence. Two more ban a practice.

Commentary often describes three or four states as mandating an algorithmic-pricing disclosure. On the enacted text, two do, and their sentences differ. Two others chose a prohibition instead, with no wording at all.

New York's sentence survived a First Amendment challenge in the district court (National Retail Federation v. James, No. 1:25-cv-05500, S.D.N.Y., Oct. 8, 2025). The appeal, 2d Cir. No. 25-2818, is briefed and undecided.

One widely repeated error, and how to check it in three clicks

Maryland HB 895 as introduced carried a second section, Com. Law § 13-322, that would have required every merchant pricing on personal data to display THIS PRICE WAS SET BY AN ALGORITHM OR BY USING YOUR PERSONAL DATA. That section was struck by amendment before the bill passed. It is not law. At least one published law-firm alert describes it as enacted, and so do the AI-generated answers of the major search engines; most firm alerts got it right.

The error is easy to make. The chapter law prints the struck section in strikethrough, and text extraction from the PDF silently drops the strikethrough, so the deleted section reads like enacted text. The State's own codified-statute service settles it:

Why it matters in practice: a merchant relying on a circulating summary may be building a disclosure for a Maryland section that does not exist while missing the New York section that does.

Presence is not adequacy

The New York Attorney General's letter to Maplebear Inc. (Instacart), dated January 8, 2026, addressed a page that contained the exact statutory sentence and concluded that the disclosure nonetheless failed to be clear and conspicuous, because of where and how it appeared. A compliance program that checks for the presence of a string is checking the wrong thing. The statute's standard is placement and prominence relative to the price.

The same letter is the reason a disclosure rendered as an image, placed inside a collapsed element, or split across markup so the sentence never appears as text should be treated as a risk, not as compliance.

The federal layer

On August 19, 2026 the Federal Trade Commission published a Proposed Enforcement Policy Statement Regarding Personalized Pricing (File No. P034101; docket FTC-2026-1057). Its premise is that a consumer who cannot tell a price was set from their data cannot avoid it, and that concealment is what makes the practice unavoidable. The comment period closes September 25, 2026.

A comment on the Statement was filed from this project on September 20, 2026, making the points on this page: that two states, not three or four, mandate a sentence; that the sentences differ; and that the presence of a sentence should not be mistaken for an adequate disclosure. The Commission posted it on September 21: Comment from Luker, Jason, FTC-2026-1057-1137.

What compliance looks like in the wild

A sweep in August 2026 searched for real-world instances of New York's mandated sentence. It is observation with its coverage limits stated, not a statistical sample. Two findings hold up:

• One category complies: digital-subscription publishers. The Wall Street Journal, Wired, The New Yorker, the Albany Times Union and NJ.com carry the sentence on renewal pricing, sometimes with a "New York residents: please note…" preamble. Renewal and win-back pricing keyed to an account is textbook personalized algorithmic pricing, and these publishers chose to disclose rather than stop.
• The disclosures live in renewal email, not on web pages. That matters for anyone measuring compliance from outside, and for any compliance team whose own review looks only at the website.

General retail, in the same sweep, was close to zero.

What Nullecho is, and what it is not

Nullecho is a free, open-source browser extension (MIT license) that blocks trackers, sends the Global Privacy Control signal, and shows each website a different, internally consistent device profile. It collects no data and has no accounts. The pricing research exists because building the tool required reading these statutes closely enough to find that the summaries disagree with the text.

The brief

Algorithmic and Personalized Pricing: A Four-State Compliance Brief puts the four enacted regimes side by side, read from the enacted and codified text rather than from summaries. Its sections:

1. Executive summary
2. The four-state matrix: citation, mechanism, mandated string, effective date, who is bound, trigger definitions, placement standard, exemptions, penalty, enforcement, private right of action, litigation
3. Disclosure versus ban: the distinction most coverage conflates
4. The federal layer
5. The California angle: a privacy statute pressed into pricing service
6. Observed compliance reality
7. A practical compliance checklist
8. Sources and verification status, including why an earlier draft of the brief was itself wrong about Maryland, and the method that caught it

The brief is free to anyone who asks for it by email. It is more useful read than sold. What costs money is the work after it: keeping it current as Connecticut, New Jersey and the FTC move, and checking what a specific page actually displays to a specific reader, every day, with receipts. If either of those is a problem you have, ask.