From On-Demand Advice to Autonomous Operations
At Amazon Accelerate 2026 in Seattle, Amazon announced a major evolution in its seller-facing artificial intelligence tools. On September 23, 2026, Mary Beth Westmoreland, Vice President of Worldwide Selling Partner Experience, revealed that Amazon's Seller Assistant—the AI helper integrated directly into Seller Central—has transformed from an on-demand chat interface into a continuous, 24/7 operational system. Rather than waiting for explicit prompts, the updated assistant monitors store activity around the clock, evaluates inventory velocity, tracks competitor price movements, and prepares background actions even when no human operator is logged in.
The release represents a significant shift in Amazon's multi-year AI roadmap for merchants. Following the initial introduction of generative seller tools under project names like Amelia in September 2024 and the transition toward agentic monitoring on Amazon Bedrock in September 2025, the 2026 package introduces four main features: persistent business memory, custom continuous automations branded as Seller Assistant workflows, an Amazon Selling Partner plugin connecting Seller Central to AWS Amazon Quick and Anthropic's Claude, and a global promotion granting 12 months of free Amazon Quick Plus access (valued at up to $720).
According to Amazon's published figures, Seller Assistant capabilities are now available to over 90% of selling partners worldwide in their native language, serving hundreds of thousands of active users. Amazon claims that merchants accept the assistant's recommendations more than 90% of the time, building its update around a clear thesis: sellers no longer want AI that merely answers queries during working hours; they want automated agents working continuously on their behalf.
Persistent Memory and the Canvas Workspace
The technical cornerstone of the expanded Seller Assistant is a persistent memory architecture designed to retain merchant context across sessions and applications. Previously, AI interfaces in Seller Central operated as isolated conversations, requiring merchants to repeatedly input brand policies, historical profit margins, restock lead times, or seasonal growth targets. Under the new system, Seller Assistant maintains an ongoing profile of each business's pricing patterns, replenishment cycles, and strategy goals.
This contextual memory remains synchronized as sellers move across different tools. Whether working inside Seller Central, utilizing AWS Amazon Quick, or conversing within Anthropic's Claude, the underlying business context follows the user. Amazon notes that suggestions sharpen over time with each interaction, and every recommendation arrives with an attached reasoning trail, allowing merchants to review the underlying logic before deciding whether to accept, modify, or reject a proposal.