Consumers Still Choose. But AI Chooses What They See.

My television recently stopped working. Three years ago, my buyer journey would probably have looked very different. I would have started with Google. I would have searched for the best televisions in my price range, opened far too many browser tabs, read reviews, visited retailer websites and tried to work out the difference between models that, at first glance, looked almost identical.

This time, I did something different …I asked AI.

I already knew some of what I wanted. I like Samsung televisions. I knew the size of my current TV, roughly what I was prepared to spend and that I needed a retailer that could deliver the new television, install it and take the old one away. Over the course of the conversation, I narrowed down the models worth considering, compared specifications and prices, and worked through whether I actually needed to upgrade—or whether I would be perfectly happy replacing what I had with something similar.

I still made the decision, but I didn't do all of the work that led to it.

And that distinction may prove to be one of the most important shifts in consumer behaviour as generative AI becomes embedded in everyday life.

Much of the conversation around AI and commerce has focused on the endgame: autonomous agents that buy our groceries, replenish household products or book our vacations without us lifting a finger. But consumers don't need to hand over their credit cards for AI to fundamentally change how they shop. They only need to hand over some of the thinking.

In Europe, 63% of consumers using AI in their shopping journey say they use it to compare brands, models, prices and reviews. Fifty-five percent use it to learn about a product or category, and 46% use it for discovery and inspiration. Yet consumers remain far more comfortable with AI suggesting options while leaving the final decision to a human than they are with handing over the purchase itself.

This is an important behavioral distinction. Humans have always found ways to reduce the mental work involved in making choices. We ask a friend who knows more than we do. We read expert reviews. We trust star ratings. We gravitate toward brands we recognize. Faced with hundreds of possible options, we look for shortcuts.

AI is becoming a remarkably efficient one, but it is also doing something more consequential. It can decide which products we see, which comparisons we make and which brands are worth investigating further.

In my case, I still chose the television. The more interesting question is: who chose the choices?

Humans Don't Want More Choice. They Want Less Work.

The appeal of AI in shopping is not difficult to understand. It solves a problem consumers have had for years: too much information.

For decades, businesses have competed by giving people more choice, more features, more content, and more ways to compare. In theory, this should make consumers better informed, but in practice, it often creates more work.

A simple purchase can now involve hundreds of reviews, dozens of near-identical products, and an endless stream of rankings, recommendations, and sponsored content. The modern consumer is not short of information. Quite the opposite … they are often drowning in it.

The truth is people rarely approach decisions by processing every available option rationally. We use shortcuts. Behavioral scientists call them heuristics, but in everyday life, they are simply ways of making a complicated choice manageable.

We buy the brand we already know. We ask someone we trust. We assume the product with 4.8 stars is probably a safer bet than the one with 3.6. We choose the bestseller. We look at the first few results and ignore most of the rest.

AI fits neatly into that pattern of human behavior. The difference is that it can now do much more of the filtering for us.

Research from McKinsey shows 63% of consumers are now using AI in their shopping journey to compare brands, models, prices, or reviews. Fifty-five percent used it to learn about a product or category, while 46% used it for discovery and inspiration.

These are not trivial tasks - they sit right in the middle of how consumers form preferences.

It helps to think about this as three levels of delegation.

First comes cognitive delegation: help me understand this category.

Then consideration delegation: tell me which options are worth looking at.

Finally comes decision delegation: choose for me.

Most public discussion of AI commerce has focused on the third stage. That is the most futuristic and, frankly, the most headline-friendly version: autonomous agents shopping, booking, and buying without human intervention.

But the first two stages may prove more important in the near term.

Consumers do not need to surrender the final decision for AI to influence the outcome. They only need to trust it enough to simplify the journey.

And that is where the behavioral shift becomes easy to miss.

A consumer who asks AI to identify the three best laptops for their needs may still spend an hour comparing those three models. They may visit the manufacturers' websites, read reviews, and check prices across retailers. From their perspective, they researched the purchase and made an informed decision.

But the field of possibilities was narrowed before that research even began. The consumer made the choice. AI helped decide what was worth choosing from.

The Consideration Set May Become the New Battleground

For brands, the most important consequence of AI-assisted shopping may be that visibility starts to mean something different.

In physical retail, companies fought for shelf space.

In e-commerce, they fought for search rankings, retailer placement, and paid visibility.

In an AI-mediated journey, the prize may be simpler and more unforgiving: being included in the answer at all.

That creates the possibility of what we might call algorithmic shelf space.

The internet dramatically expanded the number of products consumers could access. AI may do the opposite at the point of decision. It can take a market with hundreds of plausible options and reduce it to a handful that appear to fit one person's needs.

That is extraordinarily useful for consumers. It could also make the edges of the consideration set far more consequential.

If an AI tool recommends four brands, the difference between being third and fourth may matter enormously. The fourth brand may not simply rank slightly lower. It may never enter the buyer's field of view.

We are already seeing signs that AI is beginning to alter how consumers move through digital information. During the 2025 holiday season, Adobe reported a 693% year-on-year increase in traffic from generative AI platforms to US retail sites, although from a relatively small base.

At the same time, AI can reduce the need to click through at all. Pew Research Center found that when Google displayed an AI-generated summary, users clicked a traditional search result in 8% of visits, compared with 15% when no AI summary appeared. Only 1% clicked a source cited directly within the AI summary.

For brands, that creates two very different futures.

In one, AI becomes a referral engine that sends consumers directly to products and retailers.

In the other, AI becomes the place where much of the evaluation happens before the consumer ever reaches a brand-owned environment.

A consumer may form an impression of a company without seeing its website, reading its product page, or encountering the story the brand has carefully constructed around itself. Instead, its features, reputation, reviews, pricing, and perceived strengths may be compressed into a few sentences generated elsewhere.

The strategic question is no longer only whether consumers know your brand. It is whether your brand is legible enough, credible enough, and relevant enough to be surfaced when a machine is asked to solve a consumer's problem.

That may become a very different kind of competition - because in an AI-curated market, obscurity may not mean ranking low … it may mean not appearing at all.

One Technology. Four Generations. Four Different Ways of Giving Up Control.

The generational divide around AI is real, but it is more complicated than younger consumers use it and older consumers do not.

McKinsey found that 28% of Gen Z consumers use generative AI for shopping, compared with 16% of baby boomers. The gap is even wider in search behavior: 60% of Gen Z regularly use AI Overviews in traditional search, versus 29% of boomers. Edelman’s global research shows the same age gradient, with AI-platform use at 74% among 18–28-year-olds, 66% among 29–44-year-olds, 48% among 45–60-year-olds and 28% among those aged 61 and over.

But those numbers only tell us who is using the tech - they do not tell us how different generations are incorporating it into their decision-making.

For Gen Z, AI enters a shopping environment that was already shaped by algorithms. Product discovery happens through TikTok, creators, Reddit, social feeds and recommendation systems. The path to a purchase is rarely linear, and the idea that a platform might narrow the field before they begin evaluating options is not especially new. AI may simply make that process more direct.

Millennials may be the more commercially significant cohort to watch. They have high rates of AI adoption, but they are also further into the life stages where purchase decisions become more expensive and consequential. They are buying homes, managing families, choosing financial products and making larger household purchases. EY’s 2026 global research found millennials were slightly more likely than Gen Z to use AI for financial-product recommendations, at 31% versus 30%, compared with 18% of Gen X and 10% of boomers.

Gen X sits in an interesting middle ground. Many in this cohort remember shopping before the internet, then adapted to search engines, online reviews and e-commerce. Their likely pattern is not blind reliance but selective delegation: use AI to narrow the field, then verify the recommendation elsewhere. The journey may be shorter than it was before, but not fully handed over.

Boomers present a different challenge. Older consumers may be less likely to intentionally open an AI tool and ask for shopping advice, yet they are increasingly exposed to AI-generated information through search and other digital services. A consumer can read an AI Overview in Google without thinking of themselves as someone who “uses AI.”

That raises a problem for anyone trying to measure influence.

Ask a Gen Z consumer whether AI helped them shop and they may say yes.

Ask a boomer the same question and they may say no, even if an AI-generated summary helped shape the information they saw.

The real generational divide may not be between AI users and non-users.

It may be between people who know they are delegating part of the decision process and people who do not.

What If Consumers Can No Longer Tell Us What Influenced Them?

Market research has long relied on consumers to reconstruct their own decision-making.

Which brands did you consider? What influenced you? Where did you research? Why did you choose one over another?

Those questions assume people can see the path they took.

Increasingly, they may only see part of it.

A consumer might say they “researched it themselves,” even if an AI tool synthesized reviews, compared products, and narrowed the field before they began evaluating options. They may be able to explain why they chose Brand A over Brand B, while having no idea why Brand C never appeared.

That creates a new blind spot.

We may continue to measure the final stages of decision-making very well while missing the forces that shaped the field before the consumer ever entered it.

It also complicates attribution. If AI is embedded into search, recommendations, and digital platforms, consumers may not always recognize it as a source of influence at all. Older cohorts may be especially likely to underreport that exposure, not because it was absent, but because they did not consciously seek it out.

For researchers, the challenge is no longer simply to ask better questions. It may be to rethink what we are trying to measure.

Brand consideration may need to be understood as something increasingly co-produced by people and systems.

And for businesses, one question may become as important as “Why did the consumer choose us?” Did we ever make it into the set of choices they were shown in the first place?

At Kadence International, we help brands understand how consumer behavior is changing across digital channels, markets and generations—and what that means for the decisions people make. From online discovery and search behavior to consideration, trust and purchase, our research helps businesses see not just what consumers are doing, but why.

Talk to us about how evolving online behavior is reshaping your category, your customers and your path to growth.