Amazon and Google are using agentic artificial intelligence to change online retail. Both companies see AI as an important part of the shopping experience, not just an added feature.
Amazon explained how it uses generative and agentic AI to make finding and deciding on products easier. The company noted that its vast marketplace can be overwhelming for customers. With hundreds of millions of items, it can take a long time to compare. Amazon employs AI systems that analyze what customers mean rather than just focusing on keywords. These systems look at factors such as reviews, pricing trends, delivery times, return history, and browsing patterns.
Amazon offers these features through new tools. For instance, Rufus is an AI shopping assistant; Amazon Lens allows for visual searches; and Buy for Me enables AI agents to complete purchases for customers, even for items not directly sold on Amazon. These tools can summarize reviews, track prices, reorder products, and finalize transactions. Amazon stated that these features are constantly tested and improved based on customer feedback.
Google is taking a similar path with Google Cloud. In a report focused on retail, the company referred to agentic AI as systems that can reason, understand context, and take actions. Google said this helps retailers understand complex requests like style preferences or situational needs, enhancing discovery, personalization, and engagement.
Google also pointed out internal advantages for retailers. AI agents support store staff, customer service teams, and planners. They can gather information, assess options, and coordinate across systems, allowing employees to concentrate on decision-making and customer interaction. Google emphasized that being ready for these changes is essential. This readiness includes updating processes, having strong data systems, and training the workforce.
The move toward agentic AI is happening outside the United States as well. Tata Consultancy Services released a report stating that agentic AI represents the next step in retail automation. The firm mentioned that traditional AI systems have hit their limits. TCS proposed a model with several specialized agents instead of one large platform. These agents handle pricing, inventory, staffing, and supplier coordination.
TCS identified scenarios like proactive cart recovery and managing supply chain disruptions as potential use cases. In such situations, AI agents identify risks, choose responses, and improve their actions over time. The firm noted that adopting this approach requires redesigning workflows, strengthening data infrastructure, and implementing changes in stages.
Across various regions, companies view agentic AI as a fundamental shift in how retail decisions are made and carried out.
