Amazon aims to keep shoppers within its ecosystem with Rufus, reducing reliance on external search engines and AI tools, and encouraging product research to occur on Amazon.

Amazon's AI shopping assistant Rufus is projected to generate more than $10 billion in annual incremental sales, a figure highlighted by CEO Andy Jassy during the Q3 2025 earnings call.

Rufus tracks its downstream impact by measuring how interactions drive purchases across the marketplace using a seven-day rolling attribution model to capture delayed conversions.

Rufus has rapidly grown in 2025, with 250 million shoppers using it this year, 140% year-over-year growth in monthly active users, a 210% rise in interactions, and a 60% higher likelihood of completing a purchase among users.

New AI features rolled out in 2025 include Help Me Decide, expanding coverage from hundreds to millions of products, plus audio summaries and Amazon Lens for visual search.

In late October 2025, Amazon announced about 14,000 corporate layoffs framed as strategic restructuring rather than AI-driven cuts, and investor sentiment remained positive, helping after-hours stock gains.

Amazon reported Q3 2025 revenue of $180.2 billion (up 13%), with AWS up 20% to $33 billion—the strongest AWS growth since 2022.

Rufus is embedded in Amazon’s app and site and is trained on the full product catalog, reviews, Q&A content, and related web data to handle broad product comparisons and item-specific questions.

Amazon’s broader AI push includes a substantial capex plan of $125 billion for 2025-2026, the opening of Project Rainier (an $11 billion AI data center in Indiana), and investments in Anthropic and Trainium chips.

Internal planning projects Rufus could contribute over $700 million in operating profit in 2025, with a goal of $1.2 billion in profit contributions by 2027, including potential ad revenue from Rufus responses.