A consumer watchdog issued a warning about Google’s AI agent shopping protocol — Google says she’s wrong

Shortly after Google announced its new Universal Commerce Protocol for AI-powered shopping agents, a consumer economy watchdog sounded the alarm.

In a now viral post on X Nearly 400,000 views, Lindsay Owens wrote on Sunday: “Big/bad news for consumers. Google today announced how they plan to integrate shopping into their AI offerings, including search and Gemini. The plan includes ‘personalized upselling.’ That is, analyzing your chat data and using it to overcharge you.

Owens is executive director of the consumer economics think tank Groundwork Collaboration. Her concern stems from looking at Google’s roadmap, such as and also delve into some of the detailed specification documents. The roadmap includes a feature that supports ‘upselling’, allowing sellers to promote higher-priced items to AI shopping agents.

She also pointed to Google’s plans to adjust pricing for programs such as new member discounts or loyalty-based pricing Google CEO Sundar Pichai described when he announced the new protocol at the National Retail Federation conference.

After TechCrunch inquired about Owens’ allegations, Google responded publicly to X, speaking directly to TechCrunch to dismiss the validity of its concerns.

In a post on X, Google replied that“These claims about pricing are false. We strictly prohibit sellers from displaying prices on Google that are higher than what is shown on their site, period. 1/ The term ‘upselling’ is not about overcharging. It is a standard way for retailers to show additional premium product options that people may be interested in. The choice is always up to the user as to what they want to buy. 2/ ‘Instant Offers’ is a pilot that allows sellers to offer a *lower* priced deal or add additional services such as free shipping – it cannot be used to increase prices.”

Speaking separately to TechCrunch, a Google spokesperson said that Google’s Business Agent does not have functionality that allows it to change a retailer’s prices based on individual data.

Owens also pointed out that Google’s technical documents about dealing with a customer’s identity saying that: “The complexity of the scope should be hidden in the consent screen shown to the user.”

The Google spokesperson told TechCrunch that this isn’t about hiding what the user agrees to, but about consolidating actions (get, create, update, delete, cancel, complete) rather than requiring a user to agree to each one individually.

Even if Owens’ concerns about this specific protocol are moot, as Google claims, her general premise is still worth thinking about.

She warns that Big Tech-built retail agents could one day allow sellers to adjust prices based on what they think you’re willing to pay after analyzing your AI chats and shopping patterns. This is instead of charging everyone the same price. She calls it ‘monitoring prices’.

While Google says its agents can’t do such a thing now, it’s also true that Google is essentially an advertising company that serves brands and merchants. Last year, a federal court ordered Google to change some business search practices after ruling that the company engaged in anticompetitive conduct.

While many of us are excited to welcome a world where we have a team of AI agents doing tedious tasks for us (rescheduling doctor appointments, researching replacement mini-blinds), it doesn’t take a psychic to see what kinds of abuse are possible.

The problem is that the big tech companies that are best positioned to build agentic shopping tools also have the most mixed incentives. Their business relies on servicing the merchants and collecting data about consumers.

That means AI-powered shopping could be a big opportunity for startups building independent technology. We’re seeing the first few glimmers of AI-powered possibilities. Startups like it Dupewhich uses natural language searches to help people find affordable furniture, and Benithat uses images and text for thrifty fashion, are early entrants into this field.

Until then, the old adage probably applies: buyer beware.