Maryland Bans AI Grocery Price Hikes (HB 895)
Maryland’s HB 895 bans AI-driven grocery price increases from October 1, 2026. Here’s what changes for retailers, delivery apps, and pricing vendors.
By AI Pricing Guru Editorial Team
AI Pricing Guru articles are maintained by the editorial workflow behind the site: daily pricing snapshots, provider source checks, and review passes for model launches, subscription limits, and billing changes.
I try to keep these guides grounded in the actual price sheet. Marketing pages change tone, but input prices, output prices, limits, and cached-token discounts are what show up in budgets.
Maryland just became the first U.S. state to ban AI-driven grocery price increases based on personal consumer data.
Governor Wes Moore signed HB 895, the Protection From Predatory Pricing Act, on April 28. The law takes effect October 1, 2026 and prohibits covered food retailers and third-party delivery providers from using individualized consumer data to raise prices for specific shoppers.
The New York Times framed the story as Maryland banning “A.I.-driven price increases in grocery stores.” That’s the right consumer-facing headline. The more precise term is surveillance pricing: using data such as location, purchase history, demographics, browsing behavior, or other personal signals to estimate how much a shopper is willing to pay.
For AI pricing vendors, grocery chains, delivery platforms, and compliance teams, this is a meaningful line in the sand. Dynamic pricing isn’t dead. But personalized price hikes on essential food are now legally risky in Maryland, and other states are watching.
What Changed
Maryland’s new law bars covered food retailers and third-party delivery service providers from using “dynamic pricing” as defined in the bill: setting a personalized price for a good or service based on a consumer’s personal data.
The law applies to:
- Large food retailers with substantial grocery operations
- Third-party delivery services that sell or deliver covered food items
- Tax-exempt food items under Maryland’s sales tax rules
- Price increases based on personal data, not ordinary market-wide price changes
The law also treats violations as unfair, abusive, or deceptive trade practices, making them subject to enforcement and penalties under Maryland consumer protection law.
The official bill history shows the measure passed the Senate 41-1, cleared the House concurrence vote 100-31, and was approved by the governor as Chapter 154 on April 28.
What Does Not Change
This isn’t a blanket ban on every kind of price optimization.
Maryland’s law still allows normal retail pricing practices such as:
- Temporary discounts
- Promotional offers
- Loyalty program benefits
- Price changes tied to supply, demand, perishability, seasonality, or geography
- Corrections for pricing errors or network outages
That distinction matters. A grocery chain can still mark strawberries down when inventory is about to spoil. A delivery app can still show a coupon. A store can still run a loyalty-card discount.
What it can’t do is use personal data to decide that you, specifically, should pay more than another shopper for the same covered food item.
Pricing Impact: Before vs After Maryland HB 895
| Pricing practice | Before HB 895 | After October 1, 2026 in Maryland |
|---|---|---|
| Personalized food price increase using consumer data | Legally unclear / weakly regulated | Prohibited for covered retailers and delivery providers |
| Storewide dynamic price change based on inventory or seasonality | Allowed | Still allowed |
| Loyalty discount or promotion | Allowed | Still allowed, but loophole risk will be watched |
| Personalized discount | Allowed | Generally still allowed |
| Using protected-class data to deny advantages or accommodations | High legal risk | Explicitly restricted |
| Enforcement | General consumer protection tools | Treated as unfair/deceptive practice under the Act |
The business impact is asymmetric. Retailers can still optimize prices broadly, but the highest-margin version of AI pricing, estimating each shopper’s willingness to pay and lifting only that shopper’s price, is now off-limits for covered groceries in Maryland.
Why This Matters for AI Pricing Vendors
The FTC’s January 2025 surveillance pricing study found that pricing intermediaries can use a wide range of personal data to personalize prices or promotions, including location, demographics, browsing patterns, shopping history, and even interactions such as mouse movement or abandoned carts.
The FTC said the companies it reviewed worked with at least 250 clients across sectors including grocery, apparel, health and beauty, convenience stores, and hardware.
Maryland’s law turns that policy debate into operational compliance. If you sell pricing software into retail, you now need to separate three categories very clearly:
- Ordinary dynamic pricing based on supply, demand, inventory, and seasonality
- Personalized discounts that lower prices for shoppers
- Personalized price increases that use consumer data to charge a specific shopper more
The third category is where the legal risk is now concentrated.
Who Benefits
Consumers are the obvious winners, especially lower-income households that can’t easily comparison-shop for groceries across multiple apps and stores.
Retailers with transparent pricing systems also benefit. If your pricing engine is already based on inventory, geography, and store-level demand rather than individualized consumer profiles, Maryland’s law may become a trust signal.
Compliance-focused AI vendors get a new selling point. Products that can prove they don’t use personal data for individualized price increases will be easier for retailers to defend.
Who Loses
Surveillance-pricing vendors lose the most. A product pitch built around “maximize willingness-to-pay per shopper” is now harder to sell into grocery, at least in Maryland.
Delivery platforms also face pressure. They sit closest to user-level behavioral data, account history, location, basket composition, device signals, and purchase urgency, exactly the data regulators worry can be used to personalize prices.
Retailers relying on opaque algorithms face the broadest risk. If a pricing team can’t explain why two shoppers saw different grocery prices, that’s now a board-level compliance problem.
Practical Advice for Retailers and AI Teams
If you operate in Maryland or sell pricing tools to grocery customers, do three things before October 1.
1. Audit pricing inputs
Inventory, supplier cost, store location, tax, seasonality, and perishability are easier to defend. Customer-specific signals such as browsing history, account behavior, purchase urgency, demographics, or inferred income need a legal review.
2. Separate discounts from price increases
The law leaves room for discounts and loyalty programs, but advocates are already warning that loyalty exemptions could become loopholes. Don’t assume “we call it a promotion” will be enough if the system first raises a baseline price.
3. Keep an explainability record
For every automated pricing decision, retailers should be able to show whether the price was set at the store, region, SKU, or individual level. If AI is involved, log which features influenced the result.
Teams building AI cost models can use our token calculator and AI API pricing comparison to estimate the cost of running compliance reviews, price-audit agents, or customer-data minimization workflows across OpenAI, Anthropic, and Google models. For provider-specific model costs, see our OpenAI pricing, Anthropic pricing, and Google AI pricing pages.
What to Watch Next
Maryland probably won’t be the last state to act. Reporting from The Guardian and policy trackers point to activity or proposals in states including California, Colorado, Illinois, Massachusetts, New Jersey, New York, and Hawaii, though definitions vary.
That patchwork matters. One state may define “surveillance pricing” as personalized price increases based on personal data. Another may regulate algorithmic pricing more broadly. Another may only require disclosure.
For national retailers, the cheapest compliance route may be to build one stricter pricing architecture rather than maintain separate personalization rules by state.
Bottom Line
Maryland’s grocery pricing ban is narrow, but it’s still a major precedent. It doesn’t ban AI in retail pricing. It doesn’t ban promotions. It doesn’t stop storewide price changes.
It does ban the practice consumers fear most: using personal data to decide that one shopper should pay more for food than another shopper.
For AI pricing vendors, the message is clear: personalization that lowers prices may survive; personalization that raises prices for essential goods is entering a new regulatory era.
Sources: Maryland HB 895 bill history, The Guardian, WYPR, FTC surveillance pricing study, and the NYT alert source: Maryland Is First to Ban A.I.-Driven Price Increases in Grocery Stores.