AI is becoming a bigger part of how consumers search, shop, connect, and make decisions. But growing adoption doesn’t automatically bring growing trust. As AI takes on a larger role across the customer journey, retailers and marketers have to navigate where consumers are comfortable relying on it, where skepticism remains, and how their own use of AI can strengthen or undermine the trust they’ve already earned.

In a new EMARKETER guide titled "AI, authenticity, and the changing dynamics of consumer trust," our analysts and industry leaders give their perspectives on how AI is changing consumer trust, what puts that trust at risk, and what marketers can do to earn and maintain it.

Sky Canaves, EMARKETER’s principal analyst covering retail and ecommerce, told us how consumers' relationship with AI varies by product category and what assurances they need to build trust in the technology.

Q: What can brands do to be perceived as trusted sources when they appear in AI-generated answers?

A: Brands have to earn trust outside of AI, through the types of content that consumers and large language models (LLMs) both rely on. Most consumers don’t take AI recommendations at face value. Instead, they validate them against customer reviews, online discussions, and credible editorial coverage. These are among the same sources that LLMs use to build their shortlists, so they can essentially do double duty for human and machine audiences. Another step brands have to take is to optimize for LLM readability by striving for product content that is comprehensive, accurate, and consistent across channels. 

Q: How does trust in an AI recommendation compare with trust in recommendations from search or social platforms?

A: Trust in AI recommendations today occupies a middle ground between search and social. Consumers place relatively higher trust in search results for its reliability in surfacing and ranking relevant results, while social draws less trust because it’s harder for consumers to filter through the noise of the algorithms, sponsored content, and AI slop. Consumers are still grappling with a paradox of trust when it comes to AI. They use it and find it helpful, but can’t entirely trust the answers since they can be so confidently wrong. Consumers will lean on AI to narrow the field, but still have to verify when the outcomes matter.  

Q: During the consideration stage, how important is understanding why AI surfaced a particular product?

A: Demand for transparency increases when spending is at stake. The ability to see the sources, criteria, and reasoning behind a recommendation allows consumers to cross-check the references and draw their own conclusions. And to the extent that AI recommendations hold up, consumers will gain a bit more of the confidence in them that can build durable trust. 

Q: How does consumers’ willingness to rely on AI vary by product category?

A: AI adoption is additive in the sense that the more consumers use AI in one area, the more they’re likely to start using it in others. And that extends to shopping as well, where consumers turn to AI for product advice and recommendations across categories such as food and beverage and beauty, not just higher-consideration ones like consumer electronics. 

Q: What determines whether consumers trust AI to move from recommending a product to purchasing it on their behalf?

Consumers largely want to hold onto their ultimate decision-making power. Increasing willingness to delegate purchasing authority will only come with a reduction of the perceived risks associated with agentic purchasing. Consumers will need assurance that they will be made whole in the event that anything goes wrong, and they’ll be more likely to place their trust in the AI shopping assistants operated by the retailers who they have established relationships with and who have already won their trust. 

Q: What role can AI play in maintaining trust and loyalty after the purchase?

A: AI tools and features will add post-purchase value when they address customer pain points more efficiently than through other means. AI can help reduce friction in areas such as order tracking, personalized offers and reminders, returns assistance, and loyalty benefit maximization, but retailers have to strike a balance between relevance to the consumer and intrusiveness.