AI shopping is changing discovery - but not consumer trust

Much of the discussion around agentic AI in commerce has focused on one assumption: that consumers will eventually delegate purchasing decisions entirely to AI systems.

The vision is straightforward enough. A consumer tells an AI assistant they need a new winter coat, a family holiday or a replacement laptop and the agent independently handles everything from product selection to payment and fulfilment.

In this version of the future, traditional ecommerce interfaces become secondary, marketplaces lose influence and consumers gradually surrender active decision-making to software acting on their behalf.

Managing Director, PSE Consulting.

Underlying much of this vision is the assumption of "cognitive surrender" - the idea that consumers will become increasingly willing to defer judgement to AI systems rather than actively evaluating options themselves.

If AI agents can reliably identify the best products, prices and services, the argument goes, consumers may eventually see little value in conducting their own research or comparison.

The reality emerging so far appears more complicated.

Consumers are unquestionably embracing AI-assisted shopping at speed. Generative AI tools are already influencing how people discover products, compare options and navigate increasingly complex and crowded digital shopping environment.

Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!

But adoption does not necessarily equal surrender of control. Early behavioral signals suggest that while consumers are willing to use AI to reduce friction in discovery, they remain reluctant to outsource trust.

That distinction matters because it may ultimately define how agentic commerce develops.

AI is accelerating comparison, not replacing judgement

Recent international research across 4,000+ online shoppers in the UK, US, France and Germany found that consumers using AI shopping tools still rely heavily on traditional trust indicators before making purchases.

Price remains the dominant factor when evaluating AI-generated recommendations, but brand familiarity, customer reviews and marketplace reputation continue to exert significant influence over final purchase decisions.

In practice, consumers appear to be using AI as an “accelerated curation” layer rather than a replacement for purchasing judgement.

This is important because many early narratives around agentic commerce implied that AI would fundamentally weaken existing ecommerce structures. The expectation was that autonomous agents would increasingly bypass traditional digital storefronts, reduce the importance of marketplaces and make consumer loyalty less relevant.

So far, the opposite may be happening.

AI is undoubtedly changing where product discovery begins. Increasingly, consumers may start their shopping journey through conversational interfaces rather than search engines or marketplace homepages. But once products are surfaced, familiar mechanisms of trust still shape behavior. Consumers continue to validate recommendations through reviews, recognized sellers, loyalty schemes and established platforms before committing to a purchase.

The research supports this pattern: 89% of consumers say recognizing the seller's brand remains important when acting on an AI recommendation, while 92% say customer reviews continue to influence their final choice. Moreover, 68% say they would only consider purchasing from an unfamiliar seller after first checking reviews and ratings.

This creates a more evolutionary picture of AI commerce than many expected.

Why marketplaces may become more important, not less

Rather than disintermediating marketplaces, AI may actually reinforce them. Large ecommerce ecosystems already possess many of the attributes AI-assisted shoppers still value most: extensive review infrastructure, strong fulfilment capabilities, recognized seller protections and trusted payment environments.

If consumers continue seeking reassurance before completing purchases, those strengths remain highly relevant.

The comparison with earlier ecommerce technologies is instructive.

Social commerce, for example, has become a major channel for product discovery over the past decade, with consumers increasingly encountering products through creators, influencers and algorithmic feeds. However, in most cases the social platform is not the merchant of record.

The transaction, fulfilment, payments and post-purchase support typically still sit with retailers, marketplaces and their payment providers. Discovery has shifted but the underlying commercial infrastructure has largely remained intact.

Trust still closes the sale

Even as ecommerce has matured over the past two decades, trust has remained one of the strongest determinants of conversion. Consumers consistently look for social proof, recognizable brands and transparent pricing before purchasing online, particularly in higher-value or unfamiliar categories.

AI does not appear to be removing those instincts. Instead, it may be amplifying their importance by increasing the speed and scale of product exposure from entities like marketplaces which provide wide choice and a trusted logistics backbone.

If AI tools surface a broader range of sellers than consumers would normally encounter, trust signals become even more important as filtering mechanisms.

This has significant implications for merchants.

In an AI-mediated environment, visibility increasingly depends on structured product data, accurate inventory information and competitive pricing. AI systems can only recommend products they can properly interpret. Merchants with inconsistent catalogues, weak metadata or fragmented pricing strategies risk becoming less visible in AI-generated recommendations altogether.

Loyalty and payments remain part of the equation

One of the more overlooked questions surrounding agentic commerce is how loyalty programs will function inside AI-driven purchasing journeys.

Consumers have spent years building relationships with retailers, airlines, payment providers and marketplaces through points, rewards and membership ecosystems. If AI agents operate outside those structures, friction emerges. The research suggests these relationships remain highly relevant.

In the UK, 36% of consumers say they would be less likely to continue using AI shopping tools if doing so meant losing access to loyalty points and rewards, indicating the existing loyalty ecosystems remain an important part of consumer decision-making.

The likelihood is that loyalty systems themselves will need to become more interoperable with AI-assisted commerce experiences rather than disappearing entirely. Consumers may accept AI recommendations, but many will still expect rewards, protections and account benefits to remain integrated into the process.

The same applies to payments. As AI-driven commerce develops, consumers are unlikely to abandon expectations around payment security, fraud protection and transaction transparency simply because the purchasing interface changes.

In fact, research found that 55% of consumers expect to increase their use of retail marketplaces as AI adoption grows, suggesting that established platforms continue to play an important role in providing the fulfilment, payment infrastructure and operational trust that underpin online commerce. Trust in the payment process itself will remain fundamental to adoption.

The future of AI commerce may be more collaborative than autonomous

This broader pattern points towards a future where AI changes the mechanics of shopping without necessarily changing the underlying psychology of buying.

Consumers are happy to delegate repetitive tasks. They are comfortable automating comparison, discovery and research. But purchasing decisions still carry emotional, financial and reputational considerations that many consumers appear unwilling to fully hand over to autonomous systems.

That may evolve over time, particularly as younger consumers become more accustomed to AI-mediated decision-making. Trust in AI systems will almost certainly increase as the technology improves and becomes more embedded in daily life. But behavioral shifts in commerce often happen more gradually than technological capability alone would suggest.

Importantly, this does not diminish the scale of change AI may bring to ecommerce.

Agentic commerce still has the potential to reshape digital advertising, product discovery, marketplace competition, payments infrastructure and merchant visibility. Entire categories of ecommerce optimization may need to be rebuilt around how AI agents interpret and prioritize products.

But the idea that AI will simply replace trust appears increasingly simplistic.

At least for now, consumers do not seem to want autonomous shopping systems that make decisions entirely on their behalf. They want systems that help them make decisions more efficiently while still preserving their ability to verify, compare and validate.

That distinction may ultimately prove critical to understanding where agentic commerce actually goes next.

We've listed the best mobile payment apps.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Managing Director, PSE Consulting.