Agentic Commerce: Why Do Shoppers Delegate Purchase Decisions to AI?

Source: RTB House · 2026-08-28

Agentic commerce is reshaping the buying journey. For consumers, it means researching, comparing, and discovering products is easier and more user-centric than ever. Consumer trust in AI is already surprisingly high compared with other channels, and with AI commerce, some shoppers are even comfortable granting transactional autonomy.

Brands need to be responsive to the new paradigm if they are to seize opportunities and avoid pitfalls. This article explores, through the lens of our own research, consumer delegation habits, key privacy and control thresholds, and actionable steps that retail leaders can take to build trust.

Agentic commerce is a commercial application of agentic AI—any system, built with Machine Learning algorithms, most often coming in the form of a large language model (LLM), and generative AI tools that assist an individual user with a specific task.

Agentic AI in retail is used across the customer buying journey. It proactively and autonomously fulfills a range of goals, from research and discovery to purchase (this is agentic commerce), to tracking shipments and handling complex returns.

Our research shows that the willingness to delegate purchasing power to an AI shopping assistant varies by generation, safeguards, and price thresholds.

We found that human oversight before a transaction is placed is the main driver. It was the leading preference for 35% of consumers, but it’s particularly important to baby boomers. We found that 44% of them would prefer to sign off before an agentic AI makes a purchase decision. Baby boomers also show a comparatively higher sensitivity to purchasing only from trusted retailers, with 35% demanding this safeguard, compared with 29% of overall respondents.

Take, for example, the level of trust that users are already placing in agentic AI. In fact, 39% of U.S. shoppers (and 47% of U.S. millennials) say AI understands their purchase needs better than friends and family. And across the surveyed geographies, leading AI tools have a markedly higher trust score than established channels such as newspapers, major media outlets, and social media.

This high level of trust even extends to a meaningful group of consumers who are ready to let AI move from advisor to buyer through autonomous AI purchasing.

Although our research revealed a subset of super-trusters—consumers who are ready for AI to buy for them, even if returns protection and a seven-day return window are not in place—for most respondents, the removal of these features led to a clear decline in trust.

The data indicate that consumers at large are more comfortable using agentic commerce when provided with key safeguards. But the most striking finding for most people surely will be that a surprisingly large number of consumers are willing to trust agentic commerce to such a degree, with or without safeguards. After all, it’s natural that we should be protective of our hard-earned money and how it is spent.

The fact that the research and discovery phase of the purchase journey is extended by AI creates both challenges and opportunities for brands seeking to gain a foothold in this stage. With an expanded research window, maintaining engagement with retargeting becomes more important. More touchpoints mean more data, much of it widely dispersed and disordered. So brands should seek out advertising partners who have the technology to make sense of and extract insights from a more complex data landscape.