Instacart uses AI to charge different prices for same product

Instacart quietly used an AI-powered pricing system to run large-scale experiments that show different prices to different shoppers for identical items from the same retailers, prompting accusations of unfairness and opacity. What happened An investigation by Consumer Reports, Groundwork Collaborative and others found that Instacart shoppers in multiple U.S. cities were shown different prices for the same products from the same store at the same time, often with several distinct price points for a single item. Across hundreds of real shopping sessions, about three-quarters of tested items showed price variation, with differences averaging around 7–13% and in some cases reaching roughly 20–23% higher for certain customers’ carts. The pricing is powered by Instacart’s AI-based tool, which runs randomised tests to see how much different shoppers will pay before they stop buying a product. Analyses estimate that, at these fluctuation levels, a typical family of four relying on Instacart could end up paying around an extra 1,200 dollars per year in grocery costs, effectively turning ordinary customers into unwitting test subjects in a large-scale pricing experiment. Why it happened Instacart and participating retailers use the AI system to segment demand and measure “price sensitivity,” aiming to raise revenue by a few percentage points per sale without, in their view, meaningfully hurting shoppers. The company says the tests are short term, randomised, limited to about ten retail partners that already mark up online prices, and do not use personal or demographic data or real-time surge-style dynamic pricing.​ However, the experiments were not clearly disclosed to shoppers at the point of sale, and price differences were invisible unless people compared screens, exposing a lack of meaningful transparency and informed consent. Weak external oversight of algorithmic pricing, combined with broad contractual leeway for platforms and retailers to adjust online prices, allowed this system to operate at scale before the public understood it was being tested on them. What it means For directly affected customers, these practices mean some people, without knowing why, consistently pay more than others for identical groceries, making budgeting harder and exacerbating existing affordability pressures, especially for lower-income households who may already be price sensitive. Because the price differences are opaque, consumers cannot easily shop around or challenge the markups, weakening normal market checks on unfair or exploitative pricing.​ At a societal level, the case highlights how AI-enabled pricing can quietly shift bargaining power from consumers to large platforms and retailers, turning everyday shopping into an algorithmic experiment with asymmetric information. It raises broader questions for regulators and policymakers about when algorithmic price discrimination crosses from ordinary experimentation into a harmful practice that demands new rules on transparency, consent, and limits on how AI can be used to set essential goods prices. System 🤖 Eversight by Instacart 🔗 Developer: Instacart Country: USA Sector: Transport/logistics Purpose: Assess customer price sensitivity Technology: Machine learning; Pricing algorithm Issue: Fairness ; Transparency

Date it happened

2022-01-01

Organisation involved

Albertsons; Costco; Kroger; Safeway; Target

Product, system or model

Eversight by Instacart

This is a record of what was reported, not a finding that anyone broke the law. If it names your organisation and you believe it is wrong, the corrections process is free and open to everyone.