Shoppers Distrust AI

A new study from the IAB and Talk Shoppe has revealed that while Artificial Intelligence (AI) is transforming how consumers discover and compare products, it is still far from being the final stop in their shopping journey. The research shows that shoppers are happy to use AI tools for recommendations, but they remain cautious and often verify information through traditional channels such as retailer websites, reviews, and online forums. This indicates that trust in AI remains limited, even as it becomes a bigger part of online shopping habits.

According to the report, AI has undeniably sped up the process of product discovery. It helps consumers narrow down options quickly, compare features, and make more informed decisions in less time. However, rather than simplifying the shopping process, it has actually introduced additional steps. Shoppers are now taking extra time to cross-check the details they receive from AI before making any purchase.

The data shows that before using AI, consumers typically went through an average of 1.6 steps to make a purchase. After incorporating AI tools, this figure increased to nearly four steps. An overwhelming 95% of participants in the study confirmed that they took additional steps to verify information before completing their session. This demonstrates that while AI speeds up research, it also encourages more validation efforts.

Retailers and marketplaces have become the top destinations for this verification process. The study found that 78% of shoppers visited a retailer or marketplace website during their AI-assisted shopping journey. Around 32% of these visits were direct clicks from AI tools. This surge in visits suggests that AI is driving traffic to retailer websites — but not necessarily finalising purchases within AI interfaces themselves.

The report also highlighted a sharp rise in traffic to retailer sites after using AI. The proportion of shoppers visiting these sites increased from 20% before AI to 50% after using it. Most shoppers went on to check critical details like prices, product variants, reviews, and stock availability. This reflects a continuing preference for first-hand confirmation from trusted retail sources.

Despite AI’s growing role, trust remains a significant issue. Only 46% of the participants said they fully trusted AI shopping recommendations. The rest expressed doubts or partial confidence, often due to discrepancies between what AI suggested and what they found on actual retailer sites. This lack of trust shows that AI still has a long way to go in becoming a fully reliable shopping companion.

The study identified several key friction points that led to trust issues. Shoppers frequently encountered missing or broken links, mismatched product specifications, incorrect pricing, and outdated availability information. Some also found that AI recommendations didn’t align with their budget or compatibility requirements. These issues caused frustration and sent many users back to traditional search engines or verified reviews to confirm what they had learned.

Property and product data inconsistencies were among the main reasons shoppers lost faith in AI-generated information. If details like size, colour options, or stock availability differ from what’s listed on a retailer’s page, it can quickly erode consumer confidence. This highlights how important it is for brands and retailers to maintain consistent, up-to-date data across all their digital channels.

The findings also show that while AI tools are helpful for product discovery, they can’t replace the reassurance that comes from human-verified content and real-world reviews. Consumers trust the experiences of other buyers more than AI suggestions, especially when it comes to high-value or technical purchases. As a result, online reviews and independent forums remain essential sources of information for final decision-making.

To address these trust gaps, experts recommend several practical steps for businesses. First, they should ensure that specifications, prices, and stock levels are fully aligned between their brand and retailer listings. This consistency helps prevent the mismatched details that cause customers to second-guess AI recommendations.

Businesses are also encouraged to build detailed product comparison and “alternatives” pages. These pages should directly answer common questions that shoppers ask AI assistants — such as feature differences, price comparisons, and compatibility. Doing so can reduce the number of times customers feel the need to double-check information elsewhere.

Another key recommendation from the report is to improve structured data. By expanding structured data fields for product specifications, variants, pricing, availability, and reviews, businesses can help AI tools deliver more accurate and trustworthy recommendations. This, in turn, can enhance customer confidence and streamline the verification process.

It is also vital for brands to pay attention to customer conversations across online communities and forums. By monitoring where shoppers go to validate information, businesses can identify recurring issues or misinformation that may need to be corrected at the source. This ongoing observation can help companies maintain credibility and trust in an evolving digital shopping landscape.

Looking forward, the report concludes that AI will continue to play a growing role in product discovery. Shoppers appreciate how it simplifies research and reduces time spent searching manually. However, their confidence still depends heavily on transparent, verified, and consistent information from retailers and brands.

In summary, while AI tools are helping consumers make faster and more informed choices, human verification remains the cornerstone of trust. Retailers that prioritise accurate data and strong transparency across their platforms are more likely to retain customers’ confidence. As AI continues to evolve, the most successful brands will be those that combine automation with authenticity, ensuring that convenience never comes at the cost of credibility.