Artificial intelligence tools designed to perform a specific autonomous function with limited human interaction (commonly referred to as “agentic AI”) are changing the operation of myriad business processes, accelerating the rate at which organizations can handle data workflows and complex decision making. As AI agents transform into full-fledged virtual assistants, organizations are finding new ways to drive value by redesigning their enterprise’s digital landscape to accommodate and augment agentic AI. However, consumers and customers have increasingly been employing agentic AI to interact with companies of all sizes, highlighting that the use of agentic AI may no longer be a one-way road from businesses to customers.
Agentic Commerce – A Long Way from Brick-and-Mortar Retail
Recent insights published by the Boston Consulting Group indicate that the market is entering an era of “agentic commerce,” where retailers are increasingly interacting with AI agents acting on behalf of individuals with not enough time or motivation, to sift through the complex and ever expanding ecosystem of online commerce.
Although retail-oriented AI agents are not yet the primary touchpoint for the most critical transactional tasks (such as applying stored payment credentials or completing purchases autonomously), this technology already has significant influence on the way consumers browse the broad suite of purchasing options available to them online. For example, according to Adobe, traffic to US retail sites from Generative AI browsers and services increased by 4,700% year-over-year in July 2025.
Additionally, according to the “shopper edit” of the fourth edition of DHL’s ecommerce trends reports, consumers are specifically interested in AI shopping features such as virtual try-on (77%), AI-powered shopping assistants (76%) and voice-enabled product search (72%). This shift towards interacting with retailers through generative AI is redefining the relationship between consumer and company, presenting new challenges for retailers competing in the ecommerce space.
How Retailers Are Using AI to Stay Ahead
According to the “business edit” of the DHL ecommerce trends report, 53% of global ecommerce businesses use AI across their platforms. This includes various forms of agentic AI. The emerging technologies that retailers are most likely to include on their ecommerce platform include AI-powered product recommendations (44%), real-time inventory updates (43%), and personalized ordering dashboards and augmented reality product visualizing (both at 32%).
Crucially, retailers are less likely to be engaging independent vendors to develop custom AI integrations for their platforms; rather, retailers are enabling AI features that are already built in to the third-party tools the organizations rely on.
Although the implementation of AI presents clear advantages for retailers willing to employ the technology, ultimately consumer sentiment may be the key driver of AI adoption in ecommerce—seven in 10 shoppers want retailers to offer AI-powered shopping features, according to the DHL report.
Pros and Cons of Agentic Commerce
As customers and businesses alike evaluate and adopt agentic AI in the ecommerce space, several benefits and drawbacks are emerging. These emerging agentic AI options, similar to other generative AI applications generally available to the public, offer increased convenience and efficiency to consumers but will continue to face scrutiny to provide accurate and unbiased results as well as around data collection and privacy practices. However, consumers may have influence over the types of agentic AI that are adopted as retailers seek to match customer preference with technological growth.
For retailers, implementing agentic AI, and generative AI more broadly, in their operations enables them to personalize the experience of their customers at an exponentially faster rate than typical methods would permit. Agentic AI can also be deployed to identify areas of a business that are inefficient by learning from the repeated patterns of a company’s operations and applying those learnings to recommended changes.
Above all, the use of agentic AI presents an opportunity to improve and enhance customer satisfaction, complementing a company’s tried and true approach with the latest in retail technology. However, any retailer seeking to implement agentic AI in these ecommerce applications would likely weigh potential barriers to adoption, including those that are typical for other generative AI use cases: financial considerations, data privacy requirements, and internal approval and ongoing governance controls, as well as those that are more pronounced for agentic AI tools, such as the difficulty in implementing, or potential omission of, the “human in the loop” as oversight in the decision-making processes.
We will continue to monitor the developing use cases of agentic AI in ecommerce, including how those use cases play out in the technology and commercial contracting arena.