Amazon’s Upgraded Seller Assistant Now Works While You Sleep
One of the key announcements coming out of the Amazon Accelerate 2026 Conference is the major upgrade of Amazon’s Seller Assistant. First launched in 2023, Seller Assistant is a generative AI “companion” that answers questions, helps sellers identify trends, and generally navigate running their Amazon business.
But the new upgrade takes this to a whole new level by adding a variety of AI-powered features that promise to make running an Amazon program significantly easier. Let’s take a look at some of the new features and how you can leverage them to boost B2B sales on Amazon.
A Seller Assistant that Remembers Your Business
The biggest change is that Seller Assistant now has persistent memory. Many AI assistants start every conversation from a blank slate, so you have to re-explain your context each time. Persistent memory means Seller Assistant retains what it has learned about your business between sessions and applies it to future conversations. For Seller Assistant, that includes pricing patterns, inventory cycles, and growth goals.
In practice, Seller Assistant’s recommendations get more personalized the longer you use it, and each one comes with its reasoning so you can evaluate, adjust, or explore alternatives. What’s more, that memory also follows you across Seller Central, Amazon Quick (Amazon Web Services’ AI tool focused on analytics and development), and Claude, so context isn’t lost when you switch tools.
Amazon's “Always On” Agentic Seller Assistant
When it was first launched, the only way to get information from Seller Assistant was to ask it questions. For example, a seller might ask the old Seller Assistant, “How is the product performing?” and it would analyze a product’s performance in relation to sales trends and competing products/listings, and provide a recommended course of action.
Amazon has since made Seller Assistant agentic, which means that you can give it a goal, and it will work out the steps and carry them out for you, with your permission. Ask it to draft a restock plan or change a product’s price, and it can reason through the task and do it.
What’s more, the new Seller Assistant saves these questions as workflows that it can carry out on its own, even when you’re logged out. For each workflow, you choose whether the Seller Assistant only makes recommendations or if it takes action, and any action waits for your approval. A pricing workflow, for example, could help you compete for the Buy Box, the “Add to Cart” placement shoppers use to buy.
Leveraging Seller Assistant inside Other AI Tools
According to Amazon, 90 percent of its sellers already use AI tools from other companies to run their businesses. Rather than asking sellers to switch, Amazon has built a plugin that links your Amazon account to another AI tool. Once connected, that tool can see your listings, real-time performance metrics, inventory levels, and sales analytics. It can also act on them, the way Seller Assistant does inside Seller Central. Access is limited to what you allow, actions need your approval, and everything is logged.
The plugin launches in two places:
Amazon Quick: Amazon Web Services’ AI assistant for analytics and development, which can create documents and spreadsheets, conduct research, and connect to your email, calendar, and files. With the plugin, you can combine Amazon data with financial, marketing, and manufacturing data, and build agents for pricing, listings, or restocking. Quick shows what it plans to do and waits for your approval.
Claude (beta): Anthropic’s AI assistant. Amazon says sellers can connect in about 60 seconds with no coding.
For more on how AI is changing B2B selling on Amazon, see our white paper, The AI Revolution in B2B Sales on Amazon.
This is Great, but a Tool Isn’t a Strategy!
There should be little doubt that this major upgrade of Seller Assistant will make running an Amazon program smarter. But while AI might be able to carry out instructions, it can’t decide how low you’re willing to price before it hurts your margins or upsets your distributors, how to handle unauthorized resellers, or where your brand should show up. A workflow with the wrong guardrails just makes the wrong decisions faster. As we’ve argued before, Amazon is not a strategy, and neither is an AI tool.
And remember that these are Amazon’s tools. While the promise of a more efficient selling operation is attractive, we believe that merchants also should keep in mind that Amazon can use the data that is provided to its AI tools to enhance its own business (e.g. improving its own margins through cost increases). Sometimes these changes may not be in your best interest.
This is where Enceiba comes in. We help B2B manufacturers and brands build the Amazon strategy that informs tools like Seller Assistant. Contact us to talk through where AI fits in your program.
