Could the companies running AI shopping agents end up profiting more from your searches than you do?

This explores whether the platforms behind AI shopping agents can come out ahead of the consumers they claim to serve. The corpus says yes, and it gives several distinct reasons. The cleanest one is formal. Platforms judge agentic search only by how much revenue it converts, while consumers also count the effort and cost of searching. So a platform can prefer agentic search well before a shopper is better off switching Why do consumers delay adopting agentic search?. The two sides' interests split at the moment of adoption, before any bias in the agent shows up.

The less obvious finding is that platforms can profit from search without making consumers worse off at all. They do it by taking value from sellers. One model shows that when a platform charges for 'inspection tokens' (the cost of an agent checking out a seller), the profit-maximizing price pushes buyers to inspect more sellers than they need to. Consumer welfare stays the same, but surplus moves from sellers to the platform and the market as a whole gets less efficient Does platform pricing of inspection tokens encourage excess search?. So the question 'who gains more, platform or consumer?' can hide a third party who loses. A related result shows that a recommendation algorithm's advantage over plain price competition comes mostly from thin markets, where it can hold back trades. As more sellers compete, that edge shrinks toward nothing Does algorithmic recommendation lose its edge as sellers compete more?.

Then there's the question of whose side the agent is on. Several voices make the same argument from different angles: whoever builds and pays for the agent sets its goals. MIT's 'Coasean Singularity' work compares platform-built agents to bowling shoes, which are rented from the house and serve the house Does agent ownership determine whose interests get optimized?. Azhar argues that an agent earning referral fees from merchants cannot be loyal to both merchant and user Can an AI agent serve both merchant and user interests fairly?. Hedges warns that ad-funded agents will lean toward paid placements however well they're engineered Do advertising-funded shopping agents favor paid product placements?. If the business model rewards steering, the platform's gain from search becomes the consumer's cost.

Cheap search does not automatically rescue consumers either. One market-design paper argues that near-free agent search, combined with flawed agents, easy-to-copy product descriptions and a flood of low-quality sellers, can leave markets worse off than before Can cheap agent search actually harm shopping markets?. Microsoft's Magentic Marketplace experiments point the same way. Frontier models reach near-optimal outcomes only when search conditions are ideal. Even then, they showed a 10–30x bias toward the first offer they received, so speed can beat quality Do LLM agents achieve fair marketplace outcomes when given ideal search?. A platform that controls ranking order can profit from that bias.

The structural stakes come from what replaces today's search ads. As agents complete purchases with zero clicks, search ad revenue shrinks Do AI shopping agents destroy search ad revenue while leaving influencer marketing intact?. Services will then compete for the agent's attention instead of the user's, which recreates ranking and placement markets aimed at agents Will agents compete for attention just like users do?. Whether consumers get a fair share may depend on whether agentic commerce ends up open (any agent can shop anywhere) or closed (proprietary agents inside walled gardens) Will agentic commerce platforms adopt open or closed models?. The takeaway: the platform-versus-consumer gap is less about how good the agent is and more about who owns it, how it's paid, and whether users can switch to a different one.