AI in Retail: How it’s Being Used, Where Trust is Falling Short, and What Retailers Can Do About It

Artificial intelligence is changing how customers shop, with AI-driven e-commerce sales estimated to be over $144 billion by 2029. Now, retailers are looking for innovative ways to embed automated solutions across the customer journey.

While consumers are willing to use AI — and some having done so already — its full value is not being realized, with a significant adoption gap putting pressure on retailers to deliver customer expectations for trust, control, and reliable execution.

Agentic Commerce Today: Where it’s Gaining Traction

Many consumers are already using AI shopping tools, but it’s still in the earlier stages of adoption — for now. Over 63 million U.S. consumers are projected to use AI tools like ChatGPT or AI assistants on retailer websites in 2026 as it becomes more embedded in the e-commerce ecosystem.

Over half of consumers report they're most likely to use AI for monitored delivery and problem-solving, with 36 percent reporting cautious interest in using AI assistants that work directly with stores or delivery systems to track orders and resolve issues.

Consumers are also interested in exploring how AI assistants can help improve their decision-making processes, providing information on delivery options or return eligibility. However, while consumers continue to express more curiosity in agentic AI, actual adoption lags.

The Agentic Adoption Gap

There is a clear gap between consumers’ attitudes toward AI and their actual adoption. Radial research found that 58 percent of consumers report openness to placing orders through an AI assistant, but only 6 percent have actually done so. Consumers aren’t fully convinced to give up control to technology to take actions on their behalf.

Even with the promise of time or cost savings, over one-third of consumers report that they would only allow AI to act for them if they're able to approve each action. Another 23 percent would still only allow AI to make suggestions on their behalf, and 21 percent of consumers wouldn’t allow AI to act on their behalf at all.

When customers do choose to leverage an AI tool, their expectations for security and privacy are even higher. Nearly half of consumers say they would only be comfortable using AI support if security and privacy protocols were strong. They also still expect the human safety net, with 42 percent of consumers only expressing comfort using AI support.

To convert hesitancy into engagement, retailers need to increase their focus and investment in meeting customer expectations and increasing trust.

Building AI Trust Through Outcomes

Brands looking to unlock the full potential of agentic commerce must earn shoppers’ trust first by understanding how comfortable they are with AI and in what settings. Developing familiarity and comfortability with this technology may take time, but there are a few key steps retailers can explore now:

1. Prioritize security and transparency. Customers expect interactions to be secure at each step with transparency into how AI is making decisions for them and using their data.
2. Reinforce accuracy. While errors do occur, brands must ensure accuracy by making the right product, pricing and support information available to their AI tools. This helps instill confidence that AI agents can manage the shopping experience.
3. Foster human control. Shoppers want to maintain control in many ways. Brands can offer preference setting features so shoppers receive the best experience for them.

Brands should focus on successful, frictionless purchasing experiences to drive adoption and build trust.

AI is not a one-size-fits-all solution, and consumers will develop trust in its capabilities to different degrees. Agentic options should be tailored to the brand’s identity to allow for better alignment with the needs of their specific customer base. To remain competitive, brands cannot force agentic offerings. Rather, they should make them available to those who are interested.

The Future of Agentic Commerce

As we enter the next phase of agentic commerce — likely with wider adoption and more use cases — brands will have to rethink their operations across the entire customer journey. Execution on fulfillment, delivery, payments, fraud protection, and post-purchase resolution will become critical factors to building and maintaining customers' trust and likeliness of adopting AI tools in the future. Each facet of the customer journey, from click to delivery, will make the agentic commerce experience smarter over time. However, brands must be able to get the order into a customer’s hands as promised at the end of the day.

Robin Gomez is senior director of product management at Radial, North America’s largest 3PL fulfillment provider also offering integrated payment, fraud detection, and omnichannel solutions to modern and enterprise brands.