Amazon retired its Rufus chatbot on May 13, 2026, replacing it with Alexa for Shopping — an AI assistant embedded directly in Amazon’s search bar that doesn’t just answer questions, it takes action.

The move marks a meaningful shift in how Amazon thinks about AI in its retail experience. Rufus launched in early 2024 and reached over 300 million users in 2025. It was a traditional chatbot: you asked questions and got answers. Alexa for Shopping is designed differently. It can compare products side by side, track prices, and schedule purchases when an item hits a target price point — acting on behalf of the customer rather than just informing them.

Amazon is keeping parts of Rufus’s functionality — its shopping recommendations and purchase history analysis will be integrated into Alexa for Shopping. But the standalone Rufus experience is going away. The new system rolls out to all US customers within a week of the announcement.

The Shift From Chatbot to Agent

The distinction between what Rufus was and what Alexa for Shopping is meant to be comes down to one word: action.

A chatbot answers. An agent acts.

Rufus could tell you whether a drill was appropriate for drywall. Alexa for Shopping can find the drill, compare three options, add the best one to your cart, and schedule delivery for a specific date. Same underlying capability in terms of product knowledge — different model in terms of what happens next.

This mirrors a pattern playing out across enterprise software. Microsoft, Salesforce, ServiceNow, and Google have all moved past the “chat with AI” model toward AI systems with tools and permissions that let them do things on a user’s behalf. The chatbot as a standalone product is fading. The agent, embedded in workflows and authorized to act, is replacing it.

For Amazon, the consolidation under the Alexa brand also serves a business purpose. Alexa+ is Amazon’s subscription AI tier, and positioning Alexa as the central AI interface for shopping ties the assistant more directly to Amazon’s commerce business and advertising revenue model.

What Customers Are Getting

Alexa for Shopping lives inside Amazon’s search bar — so it’s not a separate chat window or a tab you navigate to. It’s the primary search interface.

Users can ask conversational questions about products, get comparisons, set price alerts, and authorize Alexa to complete purchases on their behalf. The integration of Rufus’s recommendation engine means the assistant draws on your purchase history and browsing behavior to personalize responses.

The key design choice is that it’s agentic by default. The intent is not just to inform the shopping decision but to accelerate and automate as much of it as the customer is comfortable delegating.

What This Means for Business

If you’re a business with customer-facing AI, Amazon’s pivot is worth paying attention to. The signal is clear: customers are moving past wanting AI to answer their questions. They want AI to handle things.

A chatbot that answers “what’s the best booking time for a plumbing job?” is useful. An agent that books the appointment, sends the confirmation, and follows up with a reminder is a different kind of tool — one that removes friction from the customer experience entirely.

The businesses that will win with AI in customer interactions are the ones moving their AI from information-dispensing to action-taking. That shift requires clear scope (what is the agent authorized to do), good integrations (can it actually do those things), and a customer experience that builds trust rather than eroding it.

For businesses in trades, property, healthcare, hospitality, or any industry where the phone is still a primary customer channel, this is exactly the transition Enterprise DNA’s Omni Voice service is built around. Voice AI employees that don’t just answer calls — they book appointments, qualify leads, handle intake, and route the right customers to the right people.

Amazon is building this for retail. The same model applies across industries.

Book a discovery call to explore what an agentic approach would look like for your customer interactions.