Amazon’s AI 'Buy For Me' Duplicates Over 500,000 Merchant Products Without Consent
Between Christmas and New Year, Amazon’s AI 'Buy For Me' tool scoured external retailer sites and auto-listed over 500,000 products—up from 65,000 at launch—without merchant consent. Misaligned listings led to wrong shipments, refund demands and fraud flags, prompting merchants to demand opt-out controls.
1. Unauthorized Merchant Enrollments Spark Operational Risks
Beginning last December, Amazon’s experimental “Buy For Me” AI tool duplicated listings for roughly 500,000 items—up from about 65,000 at launch—and began fulfilling orders on behalf of Amazon customers without notifying the original merchants. Six small-business owners, including stationery retailer Hitchcock Paper Co. and jewelry artist Karla Hackman, reported receiving orders via anonymous “buyforme.amazon” addresses for products they never agreed to list. Inaccurate listings led to mismatched items, customer complaints and refund demands that the independent sellers had to resolve themselves. Efforts to seek support from Amazon revealed no dedicated dispute-resolution system: one merchant was told to open a $39-per-month seller account to access help. A peer-organized survey of 187 affected merchants underscores the growing frustration, as artisans warn of damage to brand reputation and intellectual property exposure.
2. AWS Innovation Undermines AI Chip Obsolescence Concerns
Investors worried that rapid generational shifts in AI hardware would leave existing cloud-compute offerings obsolete were blindsided when Amazon Web Services (AWS) rolled out its latest Inferentia and Trainium3 accelerators in December. AWS claims these in-house AI chips can deliver up to 40% cost savings on inference workloads compared with prevailing third-party GPUs, undercutting bearish forecasts that Nvidia’s current-generation silicon would be displaced. Early customer benchmarks released by AWS show 20% to 30% lower latency on common transformer models versus comparable GPU instances, bolstering expectations for sustained demand in AWS’s high-margin machine-learning segment.
3. Heavy Capex on AI Infrastructure Pressures Cash Flow but Supports Long-Term Growth
Amazon’s trailing-twelve-month capital expenditures surged from $53.97 billion in Q2 2024 to $115.90 billion in Q3 2025 as it poured resources into new data centers, AI racks and its custom Trainium chip fabrication. Over the same period, free cash flow declined from $52.97 billion to $14.78 billion. While this spending weighs on near-term liquidity, management insists the investments are integral to sustaining AWS’s 20% revenue growth—AWS generated $33 billion in Q3 2025 sales with a 34.6% operating margin—and to driving future automation gains across Amazon’s fulfillment network. Analysts note that once the current build-out stabilizes, free cash flow should rebound sharply, laying the foundation for dividend initiation or accelerated share repurchases.