Amazon CEO Andy Jassy has identified a growing threat to his company’s retail dominance and surprisingly, it isn’t a competitor like Walmart or Target. It’s the rise of general-purpose artificial intelligence platforms – what he terms “horizontal agents” like ChatGPT – that could potentially siphon customers away from Amazon’s carefully constructed shopping ecosystem. This shift in focus highlights a pivotal moment in the evolving landscape of online retail, where the battle for customer loyalty is increasingly being fought on the front lines of artificial intelligence.
Jassy’s concerns, voiced to shareholders last week, center around the potential for these broadly capable AI tools to become preferred shopping destinations, bypassing the direct relationships retailers have cultivated with their customers. While acknowledging the capabilities of these “horizontal agents” in aggregating product selection, Jassy emphasized Amazon’s strengths in delivering the four key elements shoppers prioritize: broad selection, low prices, fast delivery, and trustworthy service. The core argument is that Amazon is uniquely positioned to excel at all four, while third-party AI platforms struggle to match that comprehensive offering.
The stakes are significant. Amazon’s own AI shopping assistant, Rufus, is already projected to generate over $10 billion in incremental annual sales, according to Andy Jassy, as reported by Fortune. This figure, exceeding the total yearly revenue of many public retailers, underscores the potential financial impact of AI-powered shopping experiences. The company saw 300 million customers use Rufus in 2025, and those users were 60% more likely to complete a purchase, demonstrating the tool’s effectiveness in driving conversions. This success is fueled by Rufus operating within Amazon’s closed ecosystem, allowing for continuous data collection and model refinement based on real customer interactions.
The Rise of AI-Assisted Shopping
The trend toward AI-assisted shopping is rapidly gaining momentum. A recent study by Boston Consulting Group revealed that 20% of consumers have already utilized generative AI for shopping purposes, with an additional 28% expressing a likelihood to do so in the near future. So nearly half of all consumers are entering what BCG calls “AI-assisted discovery,” fundamentally altering how products are found and purchased. The shift represents a significant change in consumer behavior, moving away from traditional search methods and toward conversational AI interfaces.
However, Jassy isn’t dismissing the potential for collaboration with these third-party AI platforms. He stressed the need for a “value exchange” that benefits both parties and, most importantly, enhances the customer experience. His critique of current integrations centers on inaccuracies in product details and pricing provided by these “horizontal agents,” highlighting the importance of reliable information in the shopping process. He believes a collaborative approach, where AI platforms leverage retailer data for accuracy, is crucial for delivering a superior experience.
Rufus vs. Sparky: Retailers’ AI Strategies
Amazon’s Rufus isn’t alone in the race to dominate AI-powered retail. Walmart is as well making significant strides with its own generative AI assistant, Sparky, integrated within its mobile app. While details about Sparky’s performance are less publicly available than those for Rufus, the development underscores the broader industry trend toward embedding AI directly into the shopping experience. Both Rufus and Sparky aim to provide customers with more intuitive and efficient ways to find and purchase products.
The key difference lies in the control over the customer journey. Rufus operates entirely within Amazon’s ecosystem, providing the company with complete visibility from initial inquiry to final purchase. This closed-loop system allows for continuous improvement of the AI model and optimization of retail media ROI. Sparky, while also within Walmart’s app, may face different challenges in maintaining the same level of data control and integration.
The Threat of “Horizontal Agents”
The concern voiced by Jassy isn’t about AI itself, but about where customers choose to begin their shopping journey. “Horizontal agents,” like ChatGPT and Google Gemini, offer a broad range of services beyond retail, potentially positioning them as the primary interface for many consumers. If these platforms become the go-to starting point for product searches, retailers risk losing direct access to valuable customer data and the ability to influence purchasing decisions. What we have is particularly concerning given the 693% year-over-year jump in referral traffic to retailers from AI platforms during the 2025 holiday season, according to Adobe Analytics data.
Jassy’s argument rests on the idea that customers will ultimately prioritize the comprehensive shopping experience offered by retailers like Amazon, encompassing selection, price, delivery, and trust. However, the convenience and versatility of general-purpose AI platforms could prove to be a powerful draw, especially for consumers seeking a unified digital experience. The outcome of this competition will likely shape the future of online retail for years to come.
Looking Ahead
Amazon is actively working to integrate OpenAI models into Alexa, potentially further enhancing its AI capabilities, as reported by CNBC. This move suggests a broader strategy of leveraging external AI technologies while simultaneously strengthening its own in-house solutions. The company’s focus remains on delivering a seamless and personalized shopping experience that keeps customers within its ecosystem. The next major update on Amazon’s AI strategy is expected during its first-quarter earnings call in April 2026, where Jassy is likely to provide further insights into Rufus’s performance and the company’s plans for addressing the challenges posed by “horizontal agents.”
What do you think? Will retailers be able to maintain their dominance in the age of AI, or will general-purpose platforms reshape the future of online shopping? Share your thoughts in the comments below.
