Amazon’s experimental Buy For Me feature is drawing mounting criticism from independent retailers, including sellers of used and refurbished goods, who allege unauthorised listing scraping, customer data capture, and operational disruption. The pilot, which uses agentic AI to complete purchases on third party websites when products are unavailable on Amazon’s own marketplace, has triggered broader concerns across the recommerce and circular economy ecosystem about platform power, data ownership, and channel conflict. This according to a recent article published recently on Recommerce Intelligence.
Pilot design and intent
Launched in April 2025 as a limited beta for a subset of United States customers, Buy For Me appears within a Shop brand site directly section when items cannot be sourced on Amazon.com. While some off-site listings redirect users to a retailer’s own checkout via a Shop Direct button, Buy For Me enables Amazon to process the transaction end to end. Using its own large language model alongside Anthropic’s Claude, Amazon’s system executes the purchase on the retailer’s website on behalf of the customer after payment details are entered on Amazon.
Uninvited participation claims
Several small retailers state they were never invited to participate in the programme, yet found their product listings replicated or distorted on Amazon. According to sellers cited by the AIM Group, these listings included outdated inventory, non-existent products, or mismatched imagery. For recommerce operators, where accurate grading, condition transparency, and stock turnover are critical, such inaccuracies present material risk to margins and customer trust.
Operational failures reported
Multiple sellers reported receiving orders they could not fulfil correctly due to discrepancies between what customers selected on Amazon and what appeared in their own backend systems. In one case, a retailer described more than 40 mismatched orders resulting from the AI-generated purchase instructions. Others flagged suspicious transactions they feared were linked to fraudulent payment methods, prompting cancellations and further uncertainty over downstream customer service liabilities.
Customer relationship erosion
Angie Chua, owner of California-based Bobo Design Studio, characterised the issue as a long-term erosion of direct customer relationships rather than a short-term revenue concern. She argued that Amazon’s intervention breaks the direct connection that small retailers rely on for repeat business, email marketing, and lifetime value optimisation. For refurbished and secondhand sellers, where repeat trade-in and resale cycles underpin circularity, such disintermediation carries strategic consequences.
Legal pressure builds
Chua is now exploring the basis for a potential class-action lawsuit, having collected information from 177 sellers who say their listings were scraped or reproduced without consent. Many operate on Shopify-based platforms and report partial or full replication of their inventories. The growing number of participants suggests the issue extends beyond isolated incidents and may indicate systemic risks associated with agentic commerce models.
Amazon response and opt-out
Amazon has stated that Buy For Me and Shop Direct are designed to help customers discover brands not currently sold on its platform while driving incremental sales for businesses. The company says retailers can opt out at any time by emailing branddirect@amazon.com and that feedback on the programme has been positive. However, sellers affected by the pilot argue that opt-out mechanisms place the burden on small businesses after exposure has already occurred.
Implications for recommerce
For the secondary smartphone and electronics market, the controversy highlights unresolved questions around data access, channel neutrality, and AI-mediated commerce. As platforms increasingly deploy agentic systems to optimise conversion and discovery, recommerce players face heightened risk of losing control over listings, pricing signals, and customer data. The outcome of this dispute may influence how refurbished and circular economy businesses engage with large marketplaces as AI-driven intermediation accelerates.
Market

Trade-in

Repair

Refurbishing






