Generative AI in Online Shopping: Amazon Expands Generative AI Features with Shopping Assistant and Health Chatbot

A 2025 survey from Amazon Web Services found that 45% of global IT leaders named generative AI as their top spending priority for the year, surpassing cybersecurity, which was the top priority for just 30% of respondents. The study also indicated that 90% of organizations are already using generative AI tools in some capacity, with nearly half moving beyond the stage of experimentation to full integration.

In keeping with this trend, Amazon CEO Andy Jassy stated that approximately 1,000 generative AI applications are being developed across the company, representing a large share of the $100 billion designated for capital expenditures in 2025. Amazon’s substantial investment in generative AI is highlighted by its recent unveiling of “Interests AI,” an interactive shopping assistant, and “Health AI,” a health and wellness chatbot, both of which aim to expand Amazon’s services and enhance user experience. 

Ways Amazon is Using Generative AI

Instead of relying on traditional keyword-based searches, Interests AI allows users to describe products in their own words. Large language models (LLMs) can then translate this more conversational language into actionable search queries that will identify products matching the request. Amazon’s goal is to create a more natural and intuitive shopping experience that doesn’t rely heavily on brand identification or knowing specific categories for items of interest. 

Generative AI is transforming how Amazon delivers personalized shopping recommendations by analyzing customer preferences, browsing behavior, and past purchase data with greater nuance. Amazon’s generative AI interprets the full context of a query and generates a curated set of product recommendations that target both the item directly and relevant interests. By understanding the user's intent and preferences through natural language, generative AI helps surface more personalized suggestions—many of which the user may not have found through traditional keyword searches. These models can generate tailored product suggestions that feel relevant and timely, helping users discover items they may not have explicitly searched for but are likely to want—ultimately making shopping both more comprehensive, efficient, and enjoyable.

Amazon is also testing its new Health AI chatbot designed to provide users with answers to health and wellness questions, suggestions for common care options, and recommendations for relevant products. While Amazon emphasizes that Health AI is not a substitute for professional medical advice and cannot provide personalized diagnoses or treatments, it can serve as a helpful resource for general health information and guidance. Responses that have been reviewed and verified by licensed clinicians in the U.S. are marked with a "clinically verified" badge, adding a layer of trust to the information provided.

“Generative AI is going to reinvent virtually every customer experience we know and enable altogether new ones about which we’ve only fantasized.” —Andy Jassy, CEO, Amazon

The Evolving Landscape of Intuitive AI Shopping

These new Amazon tools are examples of the trend towards more sophisticated chatbots and shopping assistants that can understand nuanced language and provide contextually relevant information, adding to the variety of generative AI already in use, which includes:

• Interests AI lets users describe products in natural language, translates conversational input into actionable search results, and reduces dependence on brand names or exact keywords.
• AI Shopping Guides offers curated guides for 100+ product types, includes tips like “factors to consider” and personalized suggestions, and is accessible via search, category pages, or Amazon.com/AIShoppingGuides.
• Personalized Recommendations highlights features tailored to user preferences, uses past behavior and shopping context to refine suggestions, and its content is reviewed by an LLM for quality and relevance.
• Rufus AI Shopping Assistant answers shopping questions in real time, provides tailored product suggestions using Amazon’s catalog and reviews, and enhances product discovery through a conversational interface.
• Rule-based chatbots that follow predefined conversational paths and rely on "if-then" logic to respond to specific keywords or user inputs. They are effective for handling frequently asked questions but struggle with complex or unexpected queries.
• AI-powered chatbots that use natural language processing and machine learning to understand user intent and generate more human-like responses. Generative AI, like Interests AI, Health AI, Gemini, and ChatGPT, create novel responses based on their extensive training data.
• Hybrid chatbots that combine elements of rule-based and AI-powered approaches to create more versatile and robust conversational experiences.
• Application-oriented chatbots that are designed for specific tasks or industries, such as customer service bots, sales assistants, or information bots.

How Do Generative AI Shopping Assistants and Chatbots Differ from Q&A Sites?

Generative AI shopping assistants and health chatbots work to understand the user's intent by analyzing their input and then processing relevant information from existing training data. This approach differs significantly from well-known question-and-answer sites such as WebMD, Reddit, and Quora.

AI Chatbots offer immediate responses and can engage in a conversation with the user to clarify questions and provide follow-up suggestions or information in real-time. Q&A sites typically require the user to post a question and wait for moderators or other users to provide answers, which can take time and may lack the give-and-take of a conversation.

Q&A sites rely on user-generated content that can vary in reliability, allowing for responses with incorrect information from mistaken sources or intentionally misleading bad actors. While the intent of AI aims to provide more accurate answers based on verified sources, users may still encounter misinformation and flawed results.

Amazon’s AI chatbots can offer personalized recommendations based on a user's interests and purchase history, a feature not typically found on Q&A sites. Similarly, bots can be integrated into existing platform services, allowing for a seamless transition from information to potential solutions and product ordering. This more-tailored approach to information can create more in-depth experience compared to the broader reach of Q&A sites. 

Balancing the Risks of AI Assisted Technology

While generative AI offers numerous advantages, it also poses potential problems such as generating incorrect or misleading information, reflecting biases from training data, and raising concerns about user privacy and data security. Despite its powerful capabilities, over-reliance on these tools can hinder critical thinking and problem-solving. Its lack of human empathy and vulnerability to manipulation raise serious concerns—especially in sensitive fields like medicine and mental health. These challenges highlight the growing importance of specialized education in areas such as cyberpsychology, healthcare technology, and artificial intelligence.

Exploring Artificial Intelligence and Technology Education at Capitol Tech

At Capitol Technology University, our undergraduate and graduate programs in Artificial Intelligence—including Maryland’s first-ever Bachelor of Science in AI—as well as Healthcare Technology, Cyberpsychology, and more, equip students with the technical skills, ethical foundations, and multidisciplinary insights needed to thrive in a rapidly evolving landscape.

To discover our academic programs, contact our Admissions team or request more information.

Written by Jordan Ford

Edited by Erica Decker