Big Tech’s AI Shopping Tools Raise Stakes for ‘Surveillance Pricing’ Laws

A new push by tech companies to integrate artificial intelligence into how consumers shop online could soon challenge legislative efforts to regulate a range of data-driven pricing practices across the retail economy.

Google last week debuted new AI-powered “agentic commerce” tools for its search engine and Gemini chatbot aimed at helping retailers promote and sell products directly inside conversational interfaces. Central to the rollout is what’s known as the Universal Commerce Protocol (UCP), a new standard that lets AI agents connect directly to retailers’ platforms and pull live pricing based on inventory, discounts and loyalty data.

While Google says the plan will reduce friction for shoppers and help retailers reach customers at moments of high interest, critics warn it could make online shopping more opaque for customers and easier for businesses to target shoppers by using AI chat logs and personal data. And some consumer advocates worry these AI systems will inevitably be used to fuel algorithmic and dynamic pricing systems, which channel personal data to tailor prices or offers to individual consumers — often called “surveillance pricing."

Google’s developer guidance for the standard suggests a plan to have AI systems act on a shopper’s behalf, said Lindsay Owens, executive director of Groundwork, a progressive economic policy group. Owens warned that the architecture described in the materials — including chatbots, search, ad data and conversational context — could make personalized marketing and pricing harder for consumers to see, control and understand as shopping increasingly shifts into AI-driven interfaces.

These mechanisms, she said, will serve as the “belt and suspenders” for “some pretty invasive retail [and] shopping experiences.”

“This is the holy grail for them in a lot of ways: Being able to control commerce, retail, online commerce and online retail,” Owens told Tech Policy Press.

The tension could be especially acute in New York state, where lawmakers recently enacted a new law designed to curb practices where companies use personal data to determine what individual consumers pay. The Algorithmic Pricing Disclosure Act, which took effect in November, restricts “surveillance” pricing” practices that use browsing history, location data and inferred income to set prices. It also requires companies to disclose when personal data is used to set prices.

Efforts to regulate these pricing practices are likely to remain a focus this year as momentum builds at the state level, particularly in Democratic-led legislatures.

So far this year, legislators in at least nine states have introduced new proposals to tackle the issue, while New York lawmakers have advanced several additional measures ranging from outright bans on surveillance pricing to disclosure requirements, consumer opt-out rights and limits on the types of personal data that can be used to set prices. The efforts follow the more than 25 bills introduced in at least half of states across the United States last year aimed at regulating data-driven pricing, up from just 10 the year before, according to Consumer Reports.

In Kentucky, a bipartisan group of state lawmakers introduced the Kentucky Price Fairness Act, which would ban companies from using AI or automated systems to charge consumers individualized prices based on personal data or behavioral profiling. The goal is to ensure AI systems don't threaten the basic notions of price fairness, said Kentucky State Rep. Adam Moore, one of the Democrats co-sponsoring the bill.

"We already joke that everything costs more when it’s for a wedding," Moore told Tech Policy Press. "But what happens when an algorithm knows you’re attending a wedding and quietly raises prices on airline tickets because it’s an event you can’t miss? Or worse, a funeral. I don’t want corporations using life events or vulnerability to squeeze more money out of Kentuckians."

Even when third-party platforms or AI tools are used, Moore said the bill would keep responsibility with the business selling the product: "They don’t get to outsource accountability. The obligation to comply stays with the business selling to the consumer."

A Google spokesperson said that claims AI tools will allow algorithmic pricing are inaccurate and that the standard is meant only to ensure prices shown in AI-powered search or chat experiences match retailers’ own listings. They also noted that pilot features such as "Direct Offers" are designed to offer discounts rather than different prices.

“We strictly prohibit merchants from showing prices on Google that are higher than what’s on their site, and provide clear tools to help shoppers see if they’re getting a good deal,” they said in an emailed statement

As companies begin embedding shopping and checkout into AI-driven platforms, Google isn’t the only company facing scrutiny. A recent investigation by Consumer Reports and Groundwork uncovered Instacart’s use of dynamic pricing on grocery orders, prompting the New York attorney general’s office to send Instacart a letter earlier this month demanding more details about the tests.

It's still unclear if New York's algorithmic pricing law will apply to new types of AI commerce, such as agent-to-agent transactions that humans might not see. For example, will prices transmitted through machine-readable data be governed in the same way, or will new features create new loopholes? OpenAI’s plans to introduce ads in ChatGPT’s free-tier also creates new questions about how user conversations will be protected and how AI-generated answers will be influenced by advertiser spend and other user data.

State officials largely declined to comment on the matter. New York State Senator Rachel May (D), who sponsored the algorithmic pricing law, did not provide answers when asked about how the law applies to new features like Google's. A spokesperson for Democratic New York Attorney General Letitia James also declined to provide specifics on the law’s interpretation but referred to a recent op-ed by James calling for stronger limits on surveillance pricing.

Google’s response doesn’t fully address broader policy concerns raised by consumer advocates about how personalized offers function in practice. While it addressed questions about price-setting, it doesn’t explain when eligibility is determined by AI systems using consumer data. It also doesn’t detail the data used to decide when offers are shown, how those determinations are explained to users or how transparency requirements would apply as shopping increasingly occurs through AI-driven conversations or automated shopping agents.

Google isn’t alone in facing scrutiny, and data-driven pricing controversies predate generative AI. Delta, Kroger, Target and Staples are among the companies that have faced backlash in recent decades for testing different prices based on customer data. Amazon has also repeatedly faced criticism dating back to the 2000s when it briefly tested varying DVD prices based on customers' browsing data and website cookies. In 2024, the Federal Trade Commission launched a study on how JPMorgan, Mastercard and other companies use personal data to sell products at different prices.

Consumers are already wary about AI-driven pricing even before it’s mainstream. A recent Morning Consult survey found just 45% of US adults had heard of dynamic pricing, but 59% think AI-powered price gouging is a "major concern." A majority also said it'd be unacceptable for AI to decide prices based on their personal data including race and age. Other worries include privacy, budgeting uncertainty, AI accuracy and potential price manipulation by scalpers.

As states push to regulate “surveillance” pricing, businesses have fought back. In Colorado, DoorDash, Verizon and United Airlines were among the companies reportedly opposed to a bill last year that would have banned using personal data to set prices and wages.

In Congress, some lawmakers have begun to frame surveillance and algorithmic pricing as a consumer protection issue rather than just a niche tech concern.

Last year, Sens. Ruben Gallego (D-AZ), Kirsten Gillibrand (D-NY) and Cory Booker (D-NJ) introduced a bill to ban using “surveillance” data to charge people different prices but still allow loyalty discounts. In the House, Rep. Greg Casar (D-TX) proposed blocking algorithmic systems from setting individualized prices altogether.

Worries beyond pricing

Advocates of more stringent antitrust enforcement warn that AI commerce platforms like Google’s could enable coordinated pricing and further concentrate market power.

Lee Hepner, senior legal counsel at the American Economic Liberties Project, said AI commerce tools risk eroding the transparency, predictability and consumer choice that underpin competitive markets. Hepner said the company’s potential to control the buying interface, seller access and advertising infrastructure creates conflicts of interest similar to those related to ongoing antitrust lawsuits against Google's ad tech business.

AI commerce systems that facilitate price discrimination or coordination among large retailers could also weaken small businesses’ ability to compete, Hepner argued.

It’s like walking up to an item at a store, he said, and “instead of [seeing] a price tag, it’s pulling the lever of a slot machine and hoping you’re getting the price on the market."

Some privacy hawks think the new AI commerce concerns are overblown.

Megan Gray, an attorney and founder of GrayMatters Law & Policy, said Google’s new AI-driven upselling tool in Gemini won't be all that different from existing search advertising practices. While she criticized what she called Google’s history of weak ad disclosures, the former FTC attorney noted that it's too soon to know if the giant is doing anything egregiously deceptive.

"I’m the last person to give Google the benefit of the doubt,” Gray said. “But I also refuse to wear a tinfoil hat. Upselling is a tried-and-true business practice, even before the internet. Upselling is not the same thing as surveillance pricing.”

Disclosure: Marty Swant is a 2026 Reporter in Residence with the Omidyar Network. The Omidyar Network has provided grant funding to Tech Policy Press.