A consumer advocacy group has raised concerns regarding Google's AI agent shopping process — Google, however, disagrees with her assessment

Concerns Raised Over Google’s AI Shopping Protocol

Soon after Google unveiled its Universal Commerce Protocol, designed to enable AI-driven shopping assistants, a prominent consumer advocacy group voiced serious reservations.

In a widely shared post on X that has attracted nearly 400,000 views, Lindsay Owens commented, “This is significant and troubling news for shoppers. Google has just revealed its strategy to weave shopping into its AI products, including search and Gemini. Their approach features ‘personalized upselling’—meaning your conversations could be analyzed to charge you more.”

Owens, who leads the consumer-focused think tank Groundwork Collaborative, expressed her worries after reviewing Google’s plans and examining technical documentation. One highlighted feature is “upselling,” which would allow retailers to promote higher-priced products through AI shopping agents.

She also criticized Google’s intention to modify prices for things like introductory discounts or loyalty rewards, a concept explained by CEO Sundar Pichai during the protocol’s announcement at the National Retail Federation event.

When TechCrunch reached out for comment, Google responded both publicly on X and directly to the publication, disputing Owens’ claims.

On X, Google clarified: “These statements about pricing are incorrect. We do not allow merchants to display prices on Google that are higher than those on their own sites. The term ‘upselling’ simply refers to offering users additional premium options, not overcharging. Ultimately, users decide what to purchase. Our ‘Direct Offers’ pilot lets merchants provide lower prices or perks like free shipping, but it cannot be used to increase prices.”

Additionally, a Google representative told TechCrunch that its Business Agent lacks the capability to alter retailer pricing based on individual user data.

Disrupt 2026: Be the First to Know

Sign up for the Disrupt 2026 waitlist to secure your spot when Early Bird tickets are released. Previous Disrupt events have featured industry giants such as Google Cloud, Netflix, Microsoft, Box, Phia, a16z, ElevenLabs, Wayve, Hugging Face, Elad Gil, and Vinod Khosla, among over 250 leaders and 200+ sessions designed to accelerate your growth. Meet hundreds of startups pushing innovation in every field.

• Location: San Francisco
• Dates: October 13-15, 2026

Join the Waitlist Now

Transparency and Privacy Concerns

Owens also noted that Google’s technical documentation regarding user identity management states: “The scope complexity should be hidden in the consent screen shown to the user.”

Google’s spokesperson explained to TechCrunch that this is not about concealing information from users, but rather about grouping actions (such as get, create, update, delete, cancel, complete) together so users don’t have to approve each one individually.

Broader Implications for Consumers

Even if Google’s current protocol doesn’t pose the risks Owens fears, her broader concerns merit attention.

She warns that, in the future, tech giants could enable merchants to set prices based on what AI predicts an individual is willing to pay, using data from chats and purchase history—rather than offering the same price to everyone. Owens refers to this as “surveillance pricing.”

While Google insists its current technology does not support such practices, it’s important to remember that Google’s core business is advertising for brands and merchants. Just last year, a federal court required Google to alter several search-related business practices after finding the company had engaged in anti-competitive conduct.

Although many people look forward to AI assistants handling everyday tasks—like rescheduling appointments or researching products—it’s not hard to imagine how these tools could be misused.

The challenge is that the major tech companies best equipped to develop AI shopping agents also have conflicting interests, as their business models depend on both serving merchants and collecting consumer data.

Opportunities for Startups

This environment could open doors for startups to create independent, consumer-friendly AI shopping solutions. Early innovators include Dupe, which helps users find affordable furniture through natural language queries, and Beni, which leverages text and images to assist with secondhand fashion shopping.

For now, the classic advice remains relevant: always shop with caution.