⚠ Traffic Starvation of the WebModerate threat

A subtler danger sits beneath the AI transition: search has always depended on a vast open web of publishers, and that web survives on the traffic — and advertising — that Google sends it. If AI answers increasingly satisfy users without sending them onward — Pew measured clicks falling from 15% of searches to 8% when an AI summary appears¹ — the publishers, forums, and creators whose content feeds both the answers and the index lose the traffic that sustains them, and the ecosystem search draws from could wither at exactly the moment AI needs it most.

The threat is a slow, systemic one rather than a direct competitor. As Google and others answer questions by summarizing others' work without the click-through, the incentive to publish erodes, quality content thins, and a backlash builds — in copyright suits, in publishers blocking the crawlers, in regulation demanding payment for scraped content. A search engine that starves the web it indexes is, in the long run, undermining its own supply.

The brake on this danger is that Google has every incentive to keep the web healthy and has begun to grapple with publisher compensation and traffic-sharing, and that a great deal of valuable content will keep being made regardless. Google's own interest lies in not killing the ecosystem, and it has the resources and the motive to strike new bargains with the publishers it depends on.

A moderate, long-term worry. Traffic starvation is a genuine systemic risk to the content supply that both search and AI answers rely on, and it is stirring real legal and regulatory pushback — but Google's self-interest in a living web, and its capacity to renegotiate the terms of the bargain, make this a pressure to manage rather than a wall about to fall.