(Maryland) H.B. 895: Protection from Predatory Pricing Act

Maryland General Assembly
House Economic Matters Committee
Taylor House Office Building
6 Bladen Street
Annapolis, MD 21401

Dear Chair Valderrama and Members of the Committee,

EPIC writes in support of H.B. 895, the Protection From Predatory Pricing Act, and to offer a few amendments to further protect Marylanders from these harmful practices. We commend Governor Moore and the House co-sponsors for prioritizing this legislation. Maryland has the opportunity to enact innovative policy that protects the rights, privacy, and financial security of Maryland residents and workers, just as Maryland did in 2024 with the passage of its landmark Maryland Online Data Privacy Act. At a time when policymakers are concerned about cost-of-living issues for their constituents, the impact of practices like surveillance pricing cannot be ignored.

The Electronic Privacy Information Center (EPIC) is an independent, nonpartisan, non-profit research organization in Washington, D.C., established in 1994 to protect privacy, freedom of expression, and democratic values in the information age.[1] EPIC has advocated for strong AI, privacy, and consumer protection laws at both the state and federal level for many years.[2]

Surveillance pricing regulation is urgently needed, and Maryland should act now.

There is an urgent need for legislation like H.B. 895 to address the harms caused by companies using AI systems to set individualized prices for consumers. Retailers have long sought to charge individual consumers the highest amount they are willing to pay for a product or a service to maximize profit.[3] Until recently, companies lacked the technological means to achieve this level of price discrimination.[4] Today, the widespread availability of vast troves of personal data about consumers and advanced algorithms allows retailers to infer the prices individual consumers are willing to pay and make surveillance pricing a reality.[5]

Surveillance pricing can involve a disturbingly sensitive and varied collection of personal information. Retailers can access enormous amounts of data by collecting data firsthand from their customers and purchasing data from data brokers.[6] Data brokers gather data about consumers as they engage a wide range of activities in today’s economy—as they work, eat, shop, study, socialize, browse the internet, seek medical care, or simply move about the world.[7] Then, data brokers profile and categorize individual consumers based on the personal data collected about them, including location, purchase history, economic status, mental and physical health conditions, or specific vulnerabilities.[8] For example, consumers may be categorized as expectant mothers, older people struggling financially, people having symptoms of depression, or people interested in weight loss, among countless other intimate categories.[9]

Armed with detailed profiles of consumers, surveillance pricing algorithms can make real-time tweaks to prices.[10] A major investigation of Instacart found that the platform was conducting surreptitious pricing experiments by varying grocery prices by tens of cents, making the changes difficult for consumers to detect but resulting in increased grocery costs of $1,200 a year for an average consumer.[11] Pricing algorithms can make minute-by-minute tweaks and continuously learn from customer responses in both brick-and-mortar stores and online.[12] Businesses can significantly increase revenue from implementing surveillance pricing to the detriment of everyday consumers.

Surveillance pricing harms consumers by exploiting individual consumers’ willingness to pay more while offering lower prices to other consumers. Some examples include Target charging $100 more for a TV on its app based on consumers’ location relative to a Target store, [13] Orbitz charging Mac users more when booking hotels, [14] and online booking sites charging a difference of more than $500 for the same hotel room based on the consumer’s location.[15] The price changes are divorced from the quality of the product or service and market competition, and some consumers will wind up paying more simply because the business inferred they might be willing to pay more.[16]

Setting individualized prices based on personal data is unfair. Consumers expect and desire that goods or services sold to the general public in the same geographical area are sold at the same price.[17] Consumers are in an asymmetrical power relationship with companies using surveillance pricing because they often do not know the practice is happening, meaning they don’t know that their personal data is being used to take more money out of their wallets, and they are stripped of the opportunity to use that knowledge to take their business elsewhere. The betrayal of consumer expectations is clearly evidenced by the extremely negative backlash when consumers discover that businesses are engaged in surveillance pricing. When reporting revealed that Kroger may be engaged in surveillance pricing using facial recognition, the pushback from consumers and lawmakers was swift.[18] Similarly, when Delta Airlines president told an investor conference that the company’s technology can determine prices that individuals are willing to pay, negative backlash ensued, followed by backtracking by Delta Airlines.[19]

On top of the privacy harms and undermining consumer autonomy, surveillance pricing can unfairly exploit lower-income consumers into paying more.[20] One investigation into prices on Staples’ and Home Depot’s websites found that people living in lower-income areas received worse deals compared to those from higher-income areas.[21] In a time of rising cost of living and more individuals living paycheck-to-paycheck, surveillance pricing can target people who can least afford the increased cost.

H.B. 895 takes important steps to protect Marylanders from the harms of surveillance pricing.

H.B. 895 takes the important step of banning surveillance pricing in grocery stores. This is essential because surveillance pricing is extremely difficult for consumers to detect and avoid. The opacity of this practice is by design: Companies know surveillance pricing is wildly unpopular with consumers. Algorithmic price adjustments also occur surreptitiously and quickly, making them difficult to detect. A ban on surveillance pricing in grocery stores is necessary to protect consumers from this harmful and unfair practice in the place where they buy their everyday necessities. 

With amendments, H.B. 895 could give Maryland residents more meaningful protections.

With a few key amendments, H.B. 895 could provide Marylanders with stronger protections from surveillance pricing for groceries and expand those protections in other contexts that also greatly affect affordability.

H.B. 895’s focus on privacy and affordability could be strengthened by expanding beyond the grocery context.

H.B. 895’s can be strengthened by expanding the scope beyond the grocery context. The types of businesses engaging in surveillance pricing range beyond grocery stores—they include online retailers, department stores, health and beauty retailers, home goods and furnishing stores, airlines, hotel booking sites, rental car companies, and many more types of retailers.[22] We suggest scoping the bill more broadly or building onto this bill in future sessions to ensure Marylanders are protected from surveillance pricing across the economy.

H.B. 895’s protections could be strengthened and harmonized with other state laws by relying on the definition of “personal data” that already exists in Maryland law rather than defining a new “surveillance data” term.

The Maryland legislature need not reinvent the wheel by creating a new definition for “surveillance data,” risking unpredictable and inconsistent enforcement and confusion for businesses trying to comply with multiple state laws related to privacy. The Maryland Online Data Privacy Act of 2024 already contains a definition of “personal data” that should be substituted for the definition of “surveillance data” in H.B. 895. First, this definition of personal data is one that businesses in Maryland are already familiar with, allowing for predictability and consistency in compliance with the law. Second, the definition of “personal data” is one that is the same or similar to other state laws, allowing for consistency between states. Third, the definition of “personal data” is more robust, covering any information that is “linked or can be reasonably linked,” and is not limited by the means of collection, in contrast to the definition of “surveillance data” in H.B. 895. Fourth, the amendment would not weaken protections because “personal data” as already defined in Maryland is the raw material for surveillance pricing. Lastly, the definition of “surveillance data” contains terms that are undefined, such as “consumer information,” which unnecessarily creates ambiguity when a robust definition for personal data already exists in Maryland law.

H.B. 895’s current exception for discounts in the definition of “dynamic pricing” is unnecessary and risks creating a large loophole, so this exception should be struck or significantly narrowed.

Surveillance pricing techniques are not necessary to provide discounts to users, and the “discounts” enabled by surveillance pricing often are not discounts at all. Thus, H.B. 895 should be amended to remove the exception for discounts in the definition for “dynamic pricing.”

“Discounts” and “savings” offered through surveillance pricing are meaningfully different from traditional discounts in ways that make them worse for consumers.[23] A business can offer meaningless surveillance-pricing “discounts” when the business sets the list price higher than what any given consumer is likely to pay, then offers each customer a personalized “discount” that results in the maximum price that specific customer would be willing to pay. [24] For example, if the list price of a box of cereal traditionally sold for $5 is raised to $9, each customer could receive an individualized “discount,” many of which result in a price paid of far more than $5. A consumer would perceive they are getting a discount without realizing that they are not actually getting a good deal. This practice can also be used to induce purchases by appealing to consumers’ desire to jump on a “good deal.” For example, the investigation of Instacart found that while Instacart shoppers at the same grocery location saw the same sale price for a bottle of ketchup, they saw different original list prices, potentially intimating the consumer is viewing a steep, time-limited discount and inducing more purchases.[25] Even if consumers end up paying the same price for a product, surveillance pricing can artificially improve the consumer’s perception that they received a discount, misleading consumers about how much they saved.[26]

This is far different and worse for consumers than traditional discounts where customers know the list price for a product or service, the discount amount, the reason for the discount, and whether other consumers are receiving the same discount or not. These factors help a customer determine whether they are receiving a good deal. Businesses often offer discounts when they have excess inventory and lower demand, to compete with other sellers, to offer holiday sales, or for many other reasons that should be clearly communicated to consumers. Consumers are familiar with Black Friday sales, end-of-season sales, and happy hour prices for that reason. This practice helps consumers determine whether a discount is actually a good deal. By contrast, businesses use surveillance pricing to set different list and discount prices for different consumers based on inferences about what each consumer is willing to pay, without explaining the basis for the prices.[27]

Business executives’ own descriptions of the practice undercut discounts as a meaningful way to reduce costs for consumers. Using surveillance pricing to allow every consumer to save money would undermine corporations’ profit-making mandate, making such a contention illogical. Further, the consultants and executives discuss surveillance pricing strategies as revenue-increasing, rather than cost-cutting.[28] For instance, a Delta Airlines executive told investors that using surveillance pricing would yield higher revenue, before the backlash.[29] Even modern loyalty programs, which businesses tout as a way to benefit consumers, are often used to gather even more personal data to extract and sell, target hyper-personalized offers and discounts, and degrade consumer benefits over time.[30] Today, the sale and abuse of customer data can generate more profit for companies than their actual business does.[31]

Businesses can offer discounts transparently without algorithmically setting prices using personal data. Historically, businesses have offered discounts and benefits based on certain personal characteristics that customers voluntarily share with businesses, such as student discounts, senior discounts, teacher and educator discounts, or government employee and veteran discounts.[32] Some businesses offer membership programs that set out the benefits that members can expect so that consumers can assess the cost versus the benefit, and customers are not required to participate in these discount programs if they do not wish to share the relevant information with a particular business. These discounts differ from surveillance pricing in other important ways: Retailers clearly communicate the basis for these discounts, these programs are difficult to game, and these discount programs reflect broadly accepted societal views about the abilities of certain groups to pay or circumstances that warrant lower prices.[33] Surveillance pricing, on the other hand, sets different prices for individuals surreptitiously, based on criteria unknown to the consumer, and consumers have no way to refuse participation given the vast troves of personal data about every person that is available for retailers to purchase and compile.

We recommend that Section 13–321(A)(3)(II) be removed from the definition of “dynamic pricing.” There is no consumer benefit to including a wholesale exclusion of discounts, promotional offers, or loyalty program benefits from the definition of dynamic pricing. Discount programs that are not based on surveillance pricing, like those discussed above, do not result in varying prices within a business day, meaning they would not fall under this prohibition on surveillance pricing in the first place. Using a promotional offer to set lower prices in a certain time of day—which this exception in the “dynamic pricing” definition currently allows—is simply engaging in harmful surveillance pricing by another name.

We also recommend the prohibition in Section 13–321(B) be amended to read, “A FOOD RETAILER MAY NOT ENGAGE IN DYNAMIC PRICING OR USE SURVEILLANCE DATA TO SET A PRICE OR OFFER A DISCOUNTED PRICE FOR CONSUMER GOODS OR SERVICES FOR A SINGLE CONSUMER OR A GROUP OF CONSUMERS.” This change will make clear that personal data cannot be used to set individualized or group-based discounts algorithmically.

Lastly, we recommend that a definition for “discounted price” be added to read, “Discounted price means a price that is verifiably lower than the widely available and publicly disclosed bona fide market price” to address the issue of showing artificially high list prices and offering targeted, personalized discounts. This will ensure that discounts offered by retailers are truly discounts as people typically understand this term and not a “discount” a retailed offers off of a listed price that is higher than an item’s actual price to make customers feel like they are saving money.

H.B. 895 should have narrower exceptions to discounts and differential pricing that do not harm consumers.

There are several ways that prices and discounts are offered based on personal data that are uncontroversial and do not harm consumers. H.B. 895 should be amended to accommodate those scenarios.

First, H.B. 895 should allow businesses to charge different prices for products if the price differential is solely based on cost differences in providing the good or a service. For example, a business can charge higher prices to deliver a good based on the consumer’s location data if the cost to deliver the good is higher because the consumer lives further away.

Second, uncontroversial discounts such as student or senior discounts, membership discounts or discounts based on how many items the consumer has bought before (i.e. buy 10 coffees and get one free) should be exempted from the law’s coverage. While those discounts depend on personal data, they are based on publicly known conditions, are generally uncontroversial because they reflect social norms about a customer’s ability to pay or benefits they should receive, and are not applied based on surreptitious data collection and price adjustments. Currently, the law would prohibit such discounts in grocery stores, even though they do not threaten consumer privacy. We recommend the bill exempt publicly communicated discounts that are offered to all consumers or large and defined groups of consumers, that are clear about the eligibility criteria that the consumer needs to meet, and that consumers affirmatively make the choice to claim. This way, the bill can preserve discounts that help lower the costs of groceries for Maryland consumers that do not pose a threat to their privacy.

*          *          *

EPIC commends Governor Moore and Maryland lawmakers for prioritizing the privacy of their constituents and recognizing the detrimental impact that surveillance pricing has on everyday people. With a few key amendments, H.B. 895 would better protect Maryland residents from unfair practices while keeping popular discounts that don’t pose a threat to their privacy. We urge the Committee to advance this important legislation.

Thank you for the opportunity to testify today. EPIC is happy to be a resource to the Committee on these issues.

---

[1] EPIC, About EPIC, https://epic.org/about/.  

[2] See e.g., Protecting America’s Consumers: Bipartisan Legislation to Strengthen Data Privacy and Security: Hearing before the Subcomm. on Consumer Protection & Comm. of the H. Comm. on Energy & Comm., 117th Cong. (2022) (testimony of Caitriona Fitzgerald, Deputy Director, EPIC), https://epic.org/wp-content/uploads/2022/06/Testimony_Fitzgerald_CPC_2022.06.14.pdf; EPIC Testifies in Support of Maryland Bill on High-Risk AI, EPIC (Feb. 27, 2025), https://epic.org/epic-testifies-in-support-of-maryland-bill-on-high-risk-ai/.

[3] Katie J. Wells, Lindsay Owens, Angel Han & Alan Smith, Groundwork Collaborative & Consumer Reports, Same Cart, Different Price: Instacart’s Price Experiments Cost Families at Checkout 4–5 (2025), http://groundworkcollaborative.org/wp-content/uploads/2025/12/Same-Cart-Different-Price.pdf [hereinafter “Instacart Investigation”].

[4] “Price discrimination” is the practice of charging different customers different amounts for the same product or service. Price Discrimination: Robinson-Patman Violations, FTC (last accessed Feb. 12, 2026), https://www.ftc.gov/advice-guidance/competition-guidance/guide-antitrust-laws/price-discrimination-robinson-patman-violations. See Len Sherman, Will Other Companies Follow Uber’s Lead Into The Black Hole of Opaque Algorithmic Pricing?, Medium (Sept. 16, 2025), https://len-sherman.medium.com/will-other-companies-follow-ubers-lead-into-the-black-hole-of-opaque-algorithmic-pricing-d79acd9cfe35.

[5] FTC, FTC Surveillance Pricing 6(b) Study: Research Summaries, A Staff Perspective 5 (2025), https://www.ftc.gov/system/files/ftc_gov/pdf/p246202_surveillancepricing6bstudy_researchsummaries_redacted.pdf [hereinafter “FTC Study”].

[6] FTC Study at 8–9.

[7] FTC Study at 8–9; Mayu Tobin-Miyaji, EPIC, Assessing the Assessments: Maximizing the Effectiveness of Algorithmic & Privacy Risk Assessments 6–7 (2025), https://epic.org/wp-content/uploads/2025/06/Assessing-the-Assessments-Report.pdf.

[8] FTC Study at 2 n. 10, 4.

[9] Jon Keegan & Joel Eastwood, From “Heavy Purchasers” of Pregnancy Tests to the Depression-Prone: We Found 650,000 Ways Advertisers Label You, The Markup (June 8, 2023), https://themarkup.org/privacy/2023/06/08/from-heavy-purchasers-of-pregnancy-tests-to-the-depression-prone-we-found-650000-ways-advertisers-label-you.

[10] FTC Study at 3–7.

[11] Instacart Investigation at 3.

[12] Instacart-owned Eversight, which sells pricing tools, admits that shoppers will see different prices. Eversight by Instacart: AI-Powered Price Optimization, Instacart Platform (last accessed Jan. 28, 2026), https://www.instacart.com/company/retailer-platform/connected-stores/eversight.

[13] Chris Hrapsky, The Target App Price Switch: What You Need to Know, KARE (Jan. 27, 2019), https://www.kare11.com/article/money/consumer/the-target-app-price-switch-what-you-need-to-know/89-9ef4106a-895d-4522-8a00-c15cff0a0514.

[14] Dana Mattioli, On Orbitz, Mac Users Steered to Pricier Hotels, Wall Street Journal (Aug. 23, 2012), https://www.wsj.com/articles/SB10001424052702304458604577488822667325882.

[15] Keith A. Spencer, Hotel Booking Sites Show Higher Prices to Travelers from Bay Area, SFGate (last updated Feb. 3, 2025), https://www.sfgate.com/travel/article/hotel-booking-sites-overcharge-bay-area-travelers-20025145.php.

[16] AI Now Institute et al., Prohibiting Surveillance Prices and Wages 11–14 (2025), http://www.economicliberties.us/wp-content/uploads/2025/02/Real-Surveillance-Prices-and-Wages-Report.pdf.

[17] See Consumer Reports, American Experiences Survey: A Nationally Representative Multi-Mode Survey, 8 (Sept. 2025), https://article.images.consumerreports.org/image/upload/v1760040676/prod/content/dam/surveys/Consumer_Reports_AES_September_2025.pdf (A survey of 2,240 U.S. adults in 2025 found that 72 percent of people who have used Instacart in the previous year did not want the company to charge different users different prices for any reason.).

[18] Mayu Tobin-Miyaji, Kroger’s Surveillance Pricing Harms Consumers and Raises Prices, With or Without Facial Recognition, EPIC (Feb. 14, 2025), https://epic.org/krogers-surveillance-pricing-harms-consumers-and-raises-prices-with-or-without-facial-recognition/.

[19] Jay Stanley, “Surveillance Pricing” Hurts Consumers, Incentivizes More Corporate Spying on Them, ACLU (Sept. 12, 2025), https://www.aclu.org/news/privacy-technology/surveillance-pricing.

[20] Seth Frotman & Tara Mikkilineni, The Trump Administration Wants to Reboot Redlining, Jolt Digest (July 7, 2025), https://jolt.law.harvard.edu/digest/the-trump-administration-wants-to-reboot-redlining.

[21] Jennifer Valentino-DeVries, Jeremy Singer-Vine & Ashkan Soltani, Websites Vary Prices, Deals Based on Users’ Information, Wall Street Journal (last updated Dec. 12, 2012), https://www.wsj.com/articles/SB10001424127887323777204578189391813881534.

[22] FTC Study at 7; AI Now Institute et al., supra note 18 at 5; Kloczko, supra note 10, at 1–2.

[23] Ma, supra note 3.

[24] Stephanie T. Nguyen, The Next Frontier of Surveillance: Investigating Pricing Systems, Yale Journal on Regulation (Sept. 21, 2025), https://www.yalejreg.com/nc/the-next-frontier-of-surveillance-investigating-pricing-systems-by-stephanie-t-nguyen/.

[25] Id.

[26] Instacart Investigation at 12.

[27] Id. at 5.

[28] FTC Study at 10; Kravitz, supra note 20; Ma, supra note 3.

[29] Kelly McCarthy, How Delta Is Using AI for Ticket Pricing and What It Means for air travel, ABC News (Aug. 5, 2025), https://abcnews.go.com/GMA/Travel/delta-ai-ticket-pricing-means-air-travel/story?id=124343088.

[30] See Levine & Nguyen, supra note 6, at 6, 16.

[31] How Loyalty Programmes Are Keeping America’s Airlines Aloft, The Economist (Aug. 6, 2025), https://www.economist.com/business/2025/08/06/how-loyalty-programmes-are-keeping-americas-airlines-aloft.

[32] See Sherman, supra note 4.

[33] Id.