A consumer economics watchdog has warned that Google’s newly announced AI-powered shopping framework could open the door to unfair pricing practices, a claim the technology giant has firmly rejected.

The concerns surfaced shortly after Google unveiled its Universal Commerce Protocol, designed to support AI shopping agents across its products, including Search and Gemini. Lindsay Owens, executive director of the Groundwork Collaborative, sparked widespread debate after posting on social media that the new system could enable “personalized upselling” by analysing users’ chat data and charging them higher prices.

Owens said her concerns were based on Google’s published roadmap and technical specification documents, which reference features supporting upselling and price adjustments tied to factors such as loyalty programmes or new-member discounts. She argued that such capabilities could eventually allow merchants to tailor prices to individual consumers rather than offering uniform pricing.

Google has strongly disputed this interpretation. Responding publicly and in direct comments to media inquiries, the company said claims that its AI shopping tools would enable overcharging are inaccurate. Google stated that it prohibits merchants from displaying prices on its platforms that are higher than those listed on their own websites.

The company explained that “upselling” refers to the standard retail practice of suggesting premium alternatives, not increasing prices beyond what merchants already charge. It also said its “Direct Offers” pilot is designed to allow discounts or added benefits such as free shipping, and cannot be used to raise prices.

Addressing privacy concerns, Google said its business agent tools do not have the ability to alter pricing based on individual user data. It also clarified that language in its technical documents about simplifying consent screens is intended to streamline user permissions, not to obscure what users are agreeing to.

Despite Google’s reassurances, Owens warned more broadly about what she described as “surveillance pricing,” a future scenario in which AI shopping agents could analyse consumer behaviour and willingness to pay in order to customise prices. She argued that even if such practices are not possible now, they could emerge as AI tools become more sophisticated.

The debate highlights growing unease over the role of large technology companies in AI-driven commerce. While AI shopping agents promise convenience by handling tasks such as product research and price comparisons, critics note that the same companies building these tools also profit from advertising and data collection, creating potential conflicts of interest.

As AI-powered shopping continues to evolve, analysts say independent startups may find opportunities to build alternative tools that prioritise transparency and consumer trust. For now, the controversy serves as a reminder that as digital shopping becomes more automated, scrutiny over pricing, privacy and fairness is likely to intensify.