Public coverage often summarized Maryland’s 2026 effort as a sweeping “dynamic pricing” ban for groceries. The legislative record tells a more complicated story. Early versions treated dynamic pricing as time-based volatility—prices changing within a short window. Late-session revisions moved toward personalized, data-driven price increases.

The introduced House bill, HB 895 (filed Feb. 4, 2026), defined “dynamic pricing” as varying prices within a business day based on demand or other factors, including AI models recalibrating “near real-time.” The framing matched the vivid examples reported in the press—prices changing “within the same ten minutes,” as described in coverage of the proposal. That image did a lot of rhetorical work because it feels like a basic breach of the retail bargain: the shelf price should mean something.

Then the Senate narrowed. As WYPR reported when the General Assembly adjourned on April 14, 2026, the House concurred with the Senate version of the Protection From Predatory Pricing Act—and the Senate also removed House-added provisions related to broader disclosures and an electronic shelf label (ESL) ban. In the later-stage Senate text (SB 387 as engrossed/third reader), the prohibition is aimed at dynamic pricing used “to increase a price for a specific consumer,” and at using surveillance/personal data to increase a price for a consumer or group.

The effective date reported by WYPR is October 1, 2026—a concrete countdown that will shape compliance calendars, retail tech rollouts, and enforcement planning.

- Oct. 1, 2026: the reported effective date once signed (WYPR).

A plain-language way to describe early drafts: Maryland flirted with a grocery “cooling-off period.” If prices can’t change more than once in a business day, shoppers can at least trust that the number they saw at 6:10 p.m. won’t turn into a surprise by 6:25.

HB 895’s introduced definition did not simply ban all price changes. It targeted “dynamic pricing” as near-real-time variability within a business day, especially when driven by demand signals and automated decision-making. That choice matters because it’s aimed at the tempo of pricing—how quickly the number can move—not at the existence of sales, promotions, or seasonal fluctuations.

Importantly, the introduced language also carved out common and popular pricing practices. HB 895 explicitly excluded:
- Promotional pricing offers
- Loyalty program benefits
- Temporary discounts…related to retention of existing customers

Those exceptions reveal a political reality: legislators wanted to target algorithmic volatility without outlawing weekly circulars, club-card deals, or coupon-style incentives that shoppers have come to expect.

Consumer Reports, in a February 2026 advocacy write-up, characterized the bill at that stage as a ban on surveillance pricing and as limiting food price changes to no more than once every 24 hours—a clear signal that advocates saw time-based stability as part of the package, at least early on.

- HB 895 defined the targeted price variation as occurring “within a business day.” (HB 895 introduced text)

By adjournment, the legislature’s center of gravity had moved from “prices shouldn’t flicker” to “prices shouldn’t be raised on you because of what the retailer knows about you.”

In the engrossed Senate text of SB 387, the core concept is not simply dynamic pricing as time-based flexibility. The prohibition is narrower and more pointed: a food retailer or third-party delivery provider may not engage in dynamic pricing to increase a price for a specific consumer. The bill also restricts the use of surveillance/personal data to increase the price for a consumer or group.

That’s a crucial distinction for readers trying to understand what changes at the store. A rule aimed at individualized increases can still allow:
- the same storewide price change for everyone, including multiple changes in a day, depending on how the final language is interpreted and enforced;
- discounts that vary by program status (if treated as a “loyalty benefit” rather than a data-driven increase);
- promotions offered broadly.

What it tries to block is a more unsettling proposition: two shoppers seeing two different “prices” because one has been algorithmically assessed as less likely to notice, less likely to comparison shop, or more urgently in need of a product.

WYPR described “surveillance pricing” as customized prices set for individuals based on data collected about traits or characteristics, often using AI. The Federal Trade Commission has used the phrase “surveillance pricing” to discuss scenarios where the same product could have different prices or promotions depending on inputs such as consumer data, behaviors or preferences, location, time, and channel—though the agency has also noted that early staff materials include hypothetical examples due to limited public data.

- SB 387 prohibits dynamic pricing used “to increase a price for a specific consumer.” (SB 387 engrossed text)

Maryland’s debate suffered from a definitional problem that’s now baked into national conversations: people use “dynamic pricing” to mean at least two different things, and they provoke different policy responses.

In HB 895’s introduced form, dynamic pricing meant price variation within a business day based on demand or other factors, including AI recalibrating “near real-time.” Public messaging leaned hard on the “same trip” scenario—prices shifting while you’re in the store.

That kind of system is easiest to imagine with electronic shelf labels, where updates can be pushed quickly. It can also be implemented online, where prices can change between page refreshes. The fairness concern is straightforward: a shopper can’t make informed choices if the ground is moving under them.

Surveillance pricing is different in kind. The defining feature is not speed but differentiation—your price versus someone else’s, potentially driven by an inference engine trained on personal data.

SB 387’s later-stage language reads like a response to that second fear. It references “personal data” by linking to an existing Maryland definition in Commercial Law (§14–4701) in the engrossed text—an attempt to anchor enforcement to established statutory terrain rather than inventing a brand-new data taxonomy.

- The prohibition targets price increases for a “specific consumer” and relies on a statutory definition of “personal data” (SB 387 engrossed text referencing CL §14–4701).

For readers trying to translate Annapolis into aisle-level reality, the big question is simple: will groceries feel different?

A law focused on individualized increases may be largely invisible on day one. Shoppers generally don’t know whether the price they see was selected uniquely for them, especially in app-based or delivery environments where the “shelf” is a screen and pricing logic is proprietary.

Still, the law can affect behavior in at least three practical ways.

First, it can deter retailers and delivery providers from using personal data in the most aggressive way: raising the price because an algorithm predicts you’ll pay it. That’s the heart of SB 387’s “increase a price for a specific consumer” language.

Second, it can change how companies structure “discounts.” If a company wants to personalize offers, the safer route—legally and reputationally—may be to personalize downward (discounts) rather than upward (increases). The Senate language, as described in the research, is keyed to increases. That does not mean personalized discounts are always benign; it means the statutory bullseye is drawn around upward steering.

Third, it puts a spotlight on the role of third-party delivery providers as well as retailers. That matters because grocery pricing is no longer confined to the store. Many shoppers buy the same staples through an app one week and in person the next.

Retailers and their advocates have not been silent in this debate. Even when policymakers are targeting a narrow band of behavior, businesses worry about compliance ambiguity—especially when laws use terms like “dynamic pricing” that carry multiple meanings in public discourse.

From an operations perspective, grocers manage thin margins and volatile supply chains. They also run constant promotions, loyalty programs, and targeted retention offers—many of which were explicitly carved out in the introduced HB 895 text. Retailers argue that customers like lower prices and deals, and that data can help deliver them.

The fear on the business side is that a broadly read “dynamic pricing” ban could chill legitimate practices:
- adjusting prices in response to supplier costs;
- matching competitors;
- managing inventory and shrink;
- offering lawful loyalty discounts.

Those concerns help explain why the final direction appears narrower. The Senate version, as reported by WYPR, also removed House-added provisions related to disclosures and an ESL ban—suggesting lawmakers decided that the safest path was to police a specific abuse rather than rewrite grocery pricing as a whole.

A fair reading is that Maryland is trying to do something difficult: constrain a future-facing risk without banning everyday retail life. Whether SB 387 draws the line cleanly will be tested in how regulators interpret “increase,” “specific consumer,” and the use of “personal data.”

Maryland’s legislative language is technical, but the underlying disputes are easy to imagine—especially in digital commerce.

Two neighbors fill identical carts through the same grocery delivery platform. One is shown higher item prices because the system predicts they’re less price-sensitive based on browsing behavior or purchase history.

SB 387’s later-stage approach appears designed for this scenario: personalized price increases tied to personal data.

A shopper scans a QR code next to an electronic shelf label and sees $4.99. Ten minutes later, the label reads $5.49 because demand spiked.

That scenario fits the public pitch cited in coverage (“within the same ten minutes”), and it aligns with the introduced HB 895 concept of within-a-business-day variability. Whether the final enacted standard squarely addresses it depends on the final language as implemented—especially if the increase is not individualized.

A shopper without a loyalty account pays $6.00; a loyalty member pays $5.00. Is that surveillance pricing?

HB 895’s introduced text explicitly excluded “loyalty program benefits.” That carve-out suggests lawmakers did not want to criminalize the common “member price” model. The harder question is what happens when loyalty programs become deeply data-driven and begin to function as individualized steering devices rather than broad membership perks.

Laws like this live or die in definitions. Maryland’s debate illustrates that in real time. The state began the session with a concept that sounded like a bright line—no near-real-time grocery repricing—and ended closer to a discrimination rule: don’t raise prices on individuals based on personal data.

That shift raises three immediate questions.

First, how will regulators determine whether a price was increased “for a specific consumer”? Proving individualized price steering can be technically challenging. It may require audit rights, data retention, or consumer complaints with screenshots—tools that are not always visible in a headline summary.

Second, how will companies reclassify behavior? A personalized “increase” can be reframed as withholding a “discount.” If two shoppers see different prices, the system can describe the lower price as a targeted promotion rather than the higher price as a targeted penalty. The introduced HB 895 carve-outs for promotions and loyalty benefits show how much hinges on labeling.

Third, what will consumers reasonably expect? If the public thinks Maryland banned all dynamic pricing in groceries, disappointment is likely. The state’s actual direction, based on the Senate text described in the research, is more surgical: it targets a specific kind of predatory personalization.

Maryland may still shape national policy. A narrow statute can become a template if it survives real-world enforcement and avoids unintended consequences for ordinary promotions. A broader statute can become a cautionary tale if it’s unworkable. Maryland appears to have chosen the template route—whether by design or by late-session compromise.