Consumers may be considerably more comfortable handing shopping decisions to artificial intelligence than many brands assume.
A new study from global media, data and creative agency Croud finds that 69% of Americans are open to letting AI purchase products on their behalf, while three-quarters would consider using AI-powered instant checkout for at least one category.
The findings come from Croud’s latest Consumer Index, The Expansion of the Validation Economy: How to Win in a World of AI-Mediated Consumer Journeys, based on a nationally representative survey of more than 2,000 U.S. consumers.
The report points to a potentially significant change in online commerce. AI is no longer simply helping shoppers find information. Increasingly, it is becoming part of the path between a consumer’s initial question and the eventual transaction.
That could force brands to rethink a basic assumption of digital marketing: that the consumer is always the one doing the searching.
Increasingly, the consumer may be asking an AI to do it.
AI is moving upstream in the buying journey
The most important finding may not be that consumers are willing to let AI check out.
It’s that AI is influencing decisions before shoppers have necessarily settled on a brand.
Among AI users surveyed by Croud, 73% use large language models to research products or brands before making a specific decision.
That puts AI remarkably early in the consideration funnel.
Traditional search marketing has largely focused on capturing consumers after they have expressed some degree of intent. A shopper searches for running shoes, compares results and eventually clicks through to a retailer or brand.
An AI-mediated journey can work differently.
A consumer might ask an LLM for the best shoes for a particular type of running, a minimalist wardrobe for a specific aesthetic or a supplement based on a set of preferences. The AI then effectively constructs the consideration set.
The brand that never appears in that answer may never get the opportunity to compete.
That’s a significant change.
In conventional search, visibility often means ranking for a query. In AI-mediated discovery, visibility increasingly means being considered relevant enough for an AI system to recommend.
For marketers, that is a much fuzzier—and potentially much more consequential—problem.
The “validation economy” is where brands still have a shot
Croud’s research also challenges the idea that consumers will blindly accept AI recommendations.
They won’t.
Among AI users, 39% validate recommendations across four or more sources before making a purchase. More than one-third use YouTube as part of that validation process.
That creates what Croud calls the “validation economy.”
AI may narrow the field, but humans still want evidence.
Reviews, creators, communities, social platforms and brand-owned content can all play a role in deciding whether an AI-generated recommendation feels credible enough to act on.
It’s a useful distinction.
AI may become the concierge, but consumers still want to check the references.
That could make third-party credibility more important rather than less.
If an AI recommends a product, shoppers may search for reviews. They may watch a creator demonstrate it. They may check Reddit, YouTube or social platforms. They may visit the brand’s website and look for evidence supporting the recommendation.
In other words, AI can accelerate discovery without eliminating skepticism.
For marketers, that means appearing in an AI answer is only the beginning.
The harder job is making sure consumers can find enough trustworthy evidence afterward to say yes.
AI checkout is gaining ground—but not everywhere
Croud’s findings suggest consumers are also becoming more comfortable with AI taking action rather than simply offering advice.
Half of respondents say they would be comfortable allowing AI to make purchases across three or more product categories.
Routine purchases are leading the way.
Groceries and household essentials rank among the categories where consumers are most willing to trust automated checkout, which makes intuitive sense. The financial and emotional stakes are relatively low, purchases are often repetitive, and the products tend to be familiar.
If an AI knows that a household buys the same detergent every month, asking it to reorder the product is hardly a science-fiction scenario.
The more interesting question is what happens when AI moves into discretionary categories.
That’s where fashion stands out.
Fashion shoppers are already behaving differently
Croud found that AI-assisted fashion shoppers spend 56% more than non-AI users.
That is a striking difference because fashion is not an obvious category for fully automated purchasing.
Unlike groceries, clothing involves taste, fit, aesthetics and identity. Consumers often want to see multiple options and compare styles before buying.
Yet AI appears to be helping some fashion shoppers navigate that complexity.
Croud also found that AI users are 23% more likely to search by style or aesthetic rather than relying on traditional brand- or keyword-driven searches.
That suggests a shift in the way consumers express intent.
Instead of searching for a specific product, shoppers can describe the outcome they want.
Think “quiet luxury office wardrobe for a hot climate” rather than a string of brand names and product categories.
That’s a much richer query—and one that AI systems are particularly well suited to interpret.
For retailers, however, it creates a new optimization challenge.
A brand can spend years building awareness around a particular product name. But if consumers increasingly describe what they want in terms of aesthetics, use cases and personal preferences, the brand needs to be relevant to those concepts before the shopper ever types its name.
Brands may need to optimize for machines before consumers
This is where Croud’s findings intersect with the emerging discipline of AI search optimization.
Traditional SEO is built around helping search engines understand and rank content.
AI-mediated discovery adds another layer: making a brand sufficiently authoritative, relevant and well represented across the web that AI systems have reason to surface it.
The mechanics are still evolving, and there is no single equivalent of a Google ranking position for generative AI.
But the underlying principle is familiar.
Authority matters. Context matters. Consistency matters.
If a brand makes claims about a product, those claims need to be supported by credible information. If independent publishers, creators and customers consistently describe a company in particular ways, those signals can become part of the broader information environment AI systems draw from.
That makes brand building and search strategy increasingly intertwined.
Croud’s work with health and wellness company Thorne provides one example.
Thorne partnered with Croud to move from a keyword-first search strategy toward one focused on intent, context and AI visibility. According to Croud, the effort contributed to a 30% year-over-year increase in organic revenue and a 66% LLM mention rate.
Those figures are company-reported rather than an independent measurement of the broader AI-search market, but they illustrate the direction marketers are exploring.
The goal is no longer simply to rank.
It’s to be understood.
Why YouTube and creators matter more in an AI world
At first glance, AI-powered shopping might seem like bad news for creators.
If consumers ask an AI to find products, why would they watch a 12-minute product review?
Croud’s research suggests the opposite may be true.
The more consumers rely on AI to narrow their choices, the more important independent validation can become.
A creator demonstrating a product provides something a language model cannot fully replicate: firsthand experience.
A reviewer can show how a pair of shoes fits, demonstrate how a camera performs in real conditions or explain why one skincare product worked better than another.
That human layer can make an abstract AI recommendation tangible.
It also explains why creators and communities may shift further down the funnel.
For years, influencer marketing was often treated primarily as an awareness channel. In an AI-mediated shopping journey, creator content can become evidence.
That’s a more valuable—and potentially more measurable—role.
The creator isn’t simply saying, “Look at this product.”
They’re helping answer the consumer’s next question: “Should I trust this recommendation?”
Search is becoming more conversational
The fashion findings offer another clue about how search itself is changing.
Croud says AI users are 23% more likely to search by style or aesthetic.
That is important because conventional search has trained marketers to think in keywords.
AI allows consumers to think in descriptions.
The difference might seem subtle, but it changes what brands need to communicate.
A shopper doesn’t necessarily need to know the name of a style. They can describe the mood, occasion, constraints and desired result.
For example:
“I need a durable but polished travel wardrobe for a week of business meetings.”
That’s not a conventional product query.
It’s a problem statement.
AI can interpret the problem and translate it into products.
Brands that understand those underlying use cases may have an advantage over brands that simply optimize product pages around increasingly narrow keyword combinations.
This is one reason AI commerce could make marketing simultaneously more complicated and more human.
The machine handles the query.
The brand still has to understand the person behind it.
The consumer journey isn’t disappearing. It’s being rearranged.
One of the more interesting implications of Croud’s research is that AI may not eliminate traditional marketing channels.
It may redistribute their roles.
Search becomes a source of AI discovery and validation.
Brand websites become authoritative sources that provide product details and supporting information.
Creators become validators.
Reviews become evidence.
Communities become trust signals.
AI assistants become navigators and, increasingly, buyers.
That is very different from the old linear funnel.
The traditional model suggested consumers move from awareness to consideration to purchase.
The AI-mediated model looks more circular.
A consumer discovers a product through AI, validates it through creators and reviews, returns to AI for alternatives, visits the brand site for details and eventually lets an agent complete the transaction.
The consumer is still making decisions.
They’re simply outsourcing more of the information processing.
That could change what “brand awareness” means
Brand awareness has traditionally been one of the foundational goals of advertising.
But if consumers increasingly ask AI systems for recommendations without naming brands, awareness alone may not be enough.
A brand could be famous among humans and still fail to appear in an AI-generated recommendation.
Conversely, a less familiar brand with strong product information, authoritative third-party coverage and positive reviews could potentially enter the consideration set earlier.
This creates an uncomfortable possibility for established brands.
The brand name that took decades to build may not automatically guarantee a place in an AI-generated shortlist.
The machine doesn’t have to care about your Super Bowl commercial.
It needs evidence that your product fits the request.
That doesn’t make brand building obsolete. Quite the opposite.
Strong brands generate the very signals AI systems and consumers may rely on: reviews, editorial coverage, creator discussion, customer experience and authoritative content.
But the definition of brand strength could become broader.
It may increasingly include machine discoverability alongside human recognition.
Trust becomes the connective tissue
Croud’s findings ultimately point to a simple idea: consumers may be comfortable delegating tasks to AI, but they aren’t ready to delegate trust.
That distinction could shape the next stage of AI commerce.
A consumer might happily let an AI reorder household products.
But for expensive, personal or consequential purchases, consumers are likely to demand more evidence.
That’s why the validation economy matters.
AI can reduce the number of choices. It cannot necessarily remove the need for confidence.
And confidence is built through many of the same mechanisms marketers have relied on for years: credible information, social proof, expertise, reviews and memorable brand experiences.
The difference is that those assets may now need to work for two audiences simultaneously.
Humans need to understand them.
Machines need to understand them.
What marketers should watch
Croud’s findings suggest several changes worth tracking as AI commerce develops.
First, optimize for intent rather than just keywords. Consumers can now describe problems, preferences and aesthetics instead of searching for specific products.
Second, build evidence around AI recommendations. If an AI mentions a brand, consumers may immediately look for reviews, creator content and third-party validation.
Third, treat creators as conversion infrastructure. Their role may increasingly extend beyond awareness into product validation.
Fourth, make product information machine-readable and authoritative. AI systems need reliable context to understand what a product is, who it is for and why it matters.
Fifth, prepare for automated purchasing. Recurring and low-risk categories may see AI-assisted checkout first, but higher-value categories such as fashion could follow as consumer confidence grows.
And perhaps most importantly, brands need to stop treating AI search as a completely separate channel.
It’s increasingly becoming part of the customer journey itself.
The machine may make shopping faster, but people still make it meaningful
The biggest surprise in Croud’s research isn’t that consumers are warming to AI purchasing.
It’s that AI may actually make human validation more important.
When a machine produces a recommendation in seconds, the consumer’s scarce resource shifts from finding options to deciding whether those options deserve trust.
That creates an unusual dynamic.
AI could reduce the amount of human effort required to discover products while increasing the importance of the human signals that confirm a purchase.
For brands, that’s both an opportunity and a warning.
The opportunity is obvious: consumers are increasingly willing to let AI help them shop.
The warning is that being selected by an AI is not the same as being trusted by a customer.
Brands that want to win in this environment will need to be discoverable to machines, useful to consumers and credible enough to survive the validation process that comes afterward.
The next generation of commerce may therefore be less about convincing people to search for your brand and more about making sure both people and machines have a reason to recommend it.
