If AI agents start shopping and comparing for us, what happens to the ad-click money that funds publishers?

This explores what happens to businesses paid by ad-supported clicks once AI agents start reading, comparing and choosing on people's behalf. First, a gap in the collection: it says a lot about what breaks and where ad money moves. It says almost nothing about the replacement models publishers are building, such as content licensing, pay-per-crawl or subscriptions. So treat what follows as a map of the pressures on publishers, not a list of fixes.

The damage to clicks is already measurable. In a randomized field experiment with more than 1,000 users, Google's AI Overviews cut clicks to third-party sites by 39.8%. The clicks that remained weren't better and users weren't more satisfied, so publishers didn't trade volume for quality Do AI Overviews actually reduce clicks to publisher sites?. Aggregators that live on comparison traffic show the same pattern at a larger scale. Tripadvisor's hotels revenue fell 23% in one quarter. The argument is that once an agent keeps the user's demand signal and ranks options against stated criteria, an ad placed in front of a human eye stops mattering Are AI agents making attention-based ads irrelevant for aggregators?. Shopping agents compress search-to-purchase into 'zero-click' transactions, which hits search ad revenue hardest Do AI shopping agents destroy search ad revenue while leaving influencer marketing intact?.

The less obvious point is that attention competition doesn't disappear. It moves to a new audience. If people hand their goals to agents, then services compete to be picked by the agent rather than clicked by the person. That will likely bring back rankings, discovery tools and paid-influence systems built for machines Will agents compete for attention just like users do?. The ad money that leaves has a few possible destinations. One argument says it moves to social and influencer channels, because agents are good at meeting needs people already have but bad at creating new wants Do AI shopping agents destroy search ad revenue while leaving influencer marketing intact?. Another says merchants turn ad budgets into lower prices and better terms that an agent's value-maximizing logic will reward, and Google's Direct Offers pilot is an early test of this Could agents redirect merchant ad budgets toward customer savings?. Either way, a publisher that sold page position loses its product.

Whoever pays for the agent also matters. An agent funded by advertising or merchant referral fees has a built-in reason to favor paid placements, however well it's engineered Do advertising-funded shopping agents favor paid product placements?. An agent like Meta's Muse, which earns fees from merchants such as Expedia, can't be fully loyal to both the merchant and the user Can an AI agent serve both merchant and user interests fairly?. Some of that bias turns out to be surprisingly easy to adjust. One experiment found that adding a single sentence to the agent's instructions, naming the consumer as the beneficiary, removed the bias that came from the platform employing the agent Can naming the consumer fix AI agent sponsorship bias?. Agent bias can also be exploited from several sides at once. When a platform tunes its ranking and sellers tune their descriptions to the same agent weaknesses, consumer harm more than doubles, with no coordination needed Can platform ranking and seller descriptions separately harm AI shoppers?.

The takeaway: the open question isn't just how publishers replace lost clicks. It's who gets paid for shaping an agent's choice. Platforms may push agentic search before consumers have good reason to switch, because the platform judges it on conversion revenue while consumers also carry their own search costs Why do consumers delay adopting agentic search?. Whether agents stay open to many providers or get locked into closed ecosystems will decide whether independent publishers can sell to agents directly or have to go through a gatekeeper Will agentic commerce platforms adopt open or closed models?.