Instacart is charging different prices to different customers in a dangerous AI experiment, report says

A Groundwork Collaborative and Consumer Reports study of 437 shoppers in four U.S. cities found Instacart is running AI-driven price experiments that produced up to five different sales prices for the identical item at the same store and day—an average spread of 13% and instances up to 23%—with nearly three-quarters of simultaneously purchased items showing price variation; researchers say this could cost an individual shopper as much as $1,200 annually. Examples cited include Lucerne eggs at Safeway priced between $3.99 and $4.79 and Clif bars between $19.43 and $21.99; Instacart, which acquired pricing-software firm Eversight in 2022 and markets the tool as modestly revenue-enhancing for retailers, says tests are randomized, not real-time dynamic pricing, and not based on personal characteristics. The findings amplify affordability and fairness concerns for consumers, suggest potential upside to retailer/Instacart margins, and coincided with a greater-than-5% intraday decline in Instacart shares.

Analysis

A Groundwork Collaborative and Consumer Reports study of 437 shoppers in North Canton, Ohio; Saint Paul, Minn.; Washington, D.C.; and Seattle found Instacart displayed up to five different sales prices for the identical item at the same store and day, with an average spread of 13% and instances up to 23%. Cited examples include Lucerne eggs at Safeway in D.C. priced from $3.99 to $4.79 and Clif bars in Seattle priced $19.43, $19.99 and $21.99, and researchers estimate a shopper could pay as much as $1,200 more annually under the observed variation.

Instacart acquired pricing-software firm Eversight in 2022 and advertises the tool to retailers, saying tests are randomized, not based on personal characteristics and could increase store revenue by up to 3%; the company contends changes are not real-time dynamic pricing. The study counters that nearly three-quarters of simultaneously purchased items showed price variation, advocacy groups labeled the experiments "dangerous," and consumers are reported to be switching stores and brands amid high inflation; CART shares fell more than 5% intraday.

Implications include potential upside to retailer margins if the tool truly lifts revenue, but material reputational, regulatory and demand risks if consumers perceive unfair pricing; those risks could prompt regulatory scrutiny, retailer pushback or increased churn. For investors this creates higher near-term volatility for CART and raises the importance of corporate disclosure on experiment design, consumer protections and measured effects on GMV and retention.

Key Decisions for Investors

• Trim or avoid new exposure to CART until management provides detailed disclosure on the pricing experiments' selection criteria, consumer safeguards and measured impact on order volume and churn
• Monitor for formal updates on Eversight integration, any regulatory inquiries or consumer/class-action risk, retailer contract changes, and short-term GMV/order trends as triggers to reassess the position
• Consider tactical hedges or reduced position sizing given negative sentiment, the >5% intraday share decline and the risk that consumer backlash could offset any modest revenue uplift
• If management can demonstrate tests are randomized, compliant and that the claimed up-to-3% revenue benefit is realized without customer attrition, consider opportunistically rebuilding exposure