2025: The year Agentic AI broke in retail
Agentic artificial intelligence evolved from leading-edge technology to mainstream solution in retail during 2025, dominating the industry landscape.
Agentic AI builds upon the prescriptive capabilities of generative AI to streamline enterprise workflows even further. Generative AI, which is based on machine learning, can create new content and ideas and create recommendations based on analysis of volumes of data in near-real-time that were previously too big to evaluate.
[READ MORE: How agentic AI is finding its place in retail]
Agentic AI goes a step beyond by analyzing massive amounts of data in near-real-time and then automatically taking action based on the results. It was widely available at both the recent NRF 2025: Retail’s Big Show and Shoptalk 2025 retail technology conferences, but not the dominant theme.
However, at eTail Boston in August 2025, agentic AI was the featured topic of numerous sessions, with the overall assessment of speakers, panelists, exhibitors and attendees being that agentic AI has reached the early stages of being a mainstream digital commerce solution.
In the four months since, agentic AI has become a prominent feature in all types of retail solutions that streamline and manage workflows in every part of the enterprise. I previously referred to 2023 as the "year generative AI broke," comparing its explosion in retail that year to how alternative rock took over the popular music scene in 1991 (as explored in the documentary “1991: The Year Punk Broke”).
In the same vein, I’m deeming 2025 the year agentic AI broke in retail. Once again, let’s grab a plaid flannel shirt and take a look at how agentic AI permeated retail this year in three key areas:
Fully integrated shopping and checkout
One major – and rapid - development in retail agentic AI in 2025 was the emergence of complete shopping experiences conducted within agentic AI engines such as ChatGPT and Perplexity. Retailers including Walmart, Instacart and Etsy, as well as digital payment platform PayPal, are allowing shoppers to conduct natural language product searches, review personalized suggestions that may include analysis of behavioral and purchase history, and check out without ever leaving the agentic AI environment.
This evolves agentic AI from serving as a middleman that enables initial product search and then allows a customer to connect to a retailer’s site to complete a purchase to a standalone commerce platform. In some respects, it could be likened to a digital shopping mall where the retailer’s storefront is within the mall environment.
Customer service
The first splash agentic AI made in retail was as a technology to support high-functioning automated chatbots that let customers make inquiries using conversational language, and in some cases, have them resolved without any direct human interaction.
During the past year, retailers have been steadily building on that foundation to create more advanced agentic AI customer service capabilities. For example, Amazon recently expanded its Amazon Connect customer service solution with new agentic AI capabilities, such as delivering and understanding natural, human-like conversations across multiple languages and accents.
In addition, while customer service representatives talk with customers, Amazon Connect analyzes conversation context and customer sentiment to suggest next steps and also complete tasks such as preparing documentation and handling routine processes.
The agents can also combine real-time clickstream data with customer history, enabling the agents and human customer service representatives to provide real-time, personalized product suggestions before a customer makes an inquiry.
Supply chain
While much of the attention on agentic AI in retail has focused on its utility as a customer-facing solution, 2025 also saw its emergence as a powerful tool for streamlining back-end supply chain workflows.
Debenhams Group, the parent of online fashion retailers including Boohoo, BoohooMan, Debenhams, PrettyLittleThing, and Karen Millen, is rolling out a new agentic AI solution designed to help manage sales, stock and pricing more effectively across its brands. The tool combines information on stock, pricing and promotions in order to provide a clearer view of demand and performance with the goal of helping merchandisers respond quickly as shopping patterns shift through the season.
And Wakefern Food Corp. is among the first grocery retailers to deploy the Afresh Fresh Buying solution, which leverages agentic AI to orchestrate forecasting, vendor selection, truck building, and issue resolution into one seamless, intelligent workflow for fresh buyers.
As a result, the retailer hopes to help buyers prioritize daily tasks in distribution center-level operations by automating routine decisions and issue resolution to maximize service levels, freshness and margin, freeing up buyers to focus more on areas such as managing relationships with local growers and suppliers.
Meanwhile, Walmart is reengineering its entire global supply chain with real-time AI and automation, including agentic AI. These solutions support activities such as predicting demand, rerouting inventory and reducing waste.
Editor's Note: This is my last Retail Insights column of 2025. Thank you for reading my industry musings over the past year and I wish you a happy, healthy and safe holiday and New Year. Retail Insights will return in 2026 with more analysis and opinion on all things retail tech.
