You ask your shiny new AI assistant to find you a deal. It confidently spits out a recommendation, you click buy, and you feel like you’ve outsmarted the internet. But you didn’t. You just overpaid by 21.6%, and the AI smiled at you the whole time.
Google’s new AI search mode is rolling out, promising a smarter, more personalized shopping experience. But when researchers compared AI search results to traditional search, they found a glaring pattern: the AI consistently surfaces products that cost over a fifth more than the standard results.
An AI assistant that optimizes for your trust instead of your wallet isn’t a bargain hunter—it’s a digital concierge steering you to the restaurant that pays the highest kickbacks.
We’ve been conditioned to believe that AI is the ultimate truth-finder, cutting through the noise to give us the best answer. But look at what’s actually happening under the hood. When you search for a men’s Specialized cycling helmet, traditional search gives you the messy, chaotic web of resellers. You find it for £39.99, plus a £4.99 delivery fee. The AI search? It bypasses the bargain bin entirely and hands you the manufacturer’s official page at a flat £45.
Defenders will say the AI is protecting you from SEO garbage, knock-offs, and sketchy third-party sellers. And frankly, they have a point. The traditional web is flooded with refurbished items sold as new, drop-shipped junk, and keyword-stuffed scam pages. But solving that problem by defaulting to the manufacturer’s list price isn’t a stroke of genius. It’s a power transfer.
When search engines stop rewarding price competition and start rewarding brand authority, the internet stops being a marketplace and becomes an oligarchy.
This isn’t an AI accuracy problem. It’s an incentive shift. For the last twenty years, the SEO economy rewarded retailers who could undercut the competition. You won the search game by being cheaper, faster, or more aggressively optimized. But AI search creates a new gatekeeping layer. Sellers no longer need to win on price via SEO—they need to win AI citation. And who wins that game? The brands with the most authoritative, clean, official data. The manufacturers.
The result isn’t “wrong” in the sense that the AI is hallucinating a price. The result is a deliberate, mathematical bias toward authority over affordability. The 21.6% premium you’re paying is the measurable cost of the AI’s trust bias.
You aren’t paying for a better product. You’re paying a toll to the brands that figured out how to speak the AI’s language.
We thought AI would democratize information and crush prices by finding the absolute best deal in seconds. Instead, the more “helpful” the AI becomes, the more it removes the price competition that made traditional search effective for bargain hunting. It’s quietly stripping away your ability to comparison shop, all while wrapping the overcharge in a polite, conversational bow.
The next time you ask an AI to find you a product, remember what it’s actually doing. It’s not hunting for a bargain. It’s curating a reality where the cheapest option is deemed “untrustworthy” and the most expensive option is served as the default truth. The AI isn’t broken. It’s just working for the other side.
FAQ
Q: Isn't the AI just filtering out cheap, low-quality knock-offs?
A: Sometimes. But hiding a £39.99 helmet behind a £45 'trusted' listing just because the cheaper one has a £4.99 shipping fee isn't protecting you—it's just lazy math that benefits the manufacturer's bottom line.
Q: What does this mean for online retailers?
A: The traditional SEO game is dead. Sellers no longer win by undercutting competitors on price or keyword stuffing; they win by becoming the 'authoritative' source the AI decides to cite.
Q: Is this an intentional conspiracy by Google to make us spend more?
A: It's not a smoky-room conspiracy; it's an incentive shift. AI models naturally prefer clean, official data from manufacturers over messy reseller pages. The result is a 'trust premium' that just happens to make brands richer while draining your wallet.