Answering Without Referring: How AI Search Rewrites the Web's Economic Bargain

Source: econ.GN updates on arXiv.org — https://arxiv.org/abs/2607.07652 Date read: 2026-09-01 Connected to: L-001, seed-020 Kind: content Escalation: store-only Escalation rationale:

What this is

An empirical economics paper measuring the displacement of web referral traffic by AI search intermediaries (ChatGPT Search vs. Google). The core finding: ChatGPT generates outbound clicks in only 5.2% of sessions versus Google's traditional high-referral model, suggesting a structural shift in attention allocation protocols on the web.

What I took from it

This is a cleanly measured instantiation of coordination displacement (seed-020) under intermediation adoption, but it does not yet constitute a primary theoretical argument about why this happens or what generalizes. The paper establishes that the economic bargain between users, intermediaries, and content producers has shifted — the intermediary now captures value by resolving queries internally rather than routing to external sites. This fits L-001 (protocol ossification under adoption pressure): as AI search gains market share, the old referral protocol becomes locked in place by lock-in and switching costs, and new entrants or content producers cannot easily modify the bargain.

However, the paper is descriptive of the outcome (low referral rates, high internal resolution) rather than theoretically engaged with the mechanism driving it or the generalization to other protocol systems. It does not address why this particular intermediation structure emerges, under what conditions other intermediaries might adopt it, or how this pattern replicates across domains. It is a case study in the economic effects of intermediation capture, not a sustained theoretical inquiry into protocol restructuring.

Research connections

L-001: Confirms that adoption pressure on AI search intermediaries creates a stable, difficult-to-reverse shift in attention-routing protocol — the old referral model becomes harder to restore as users and the intermediary converge on internal resolution.
seed-020: Directly exemplifies symptom-hierarchy coordination displacement — the coordination norm (route users to sources) is replaced by a different norm (resolve internally) under adoption pressure, displacing the original symbiosis between intermediary and content producer.

Seed

Seed title: Internal Resolution as Protocol Irreversibility Seed type: observation Seed text: When an intermediary protocol transitions from routing/referral (low-opacity, user sends query externally) to internal resolution (high-opacity, intermediary resolves and displays answer), the economic and informational asymmetries favor lock-in of the internal model. This occurs because: (1) users experience lower friction, (2) the intermediary captures more value, and (3) content producers lose visibility into query intent. Once adoption crosses a threshold, reverting to routing becomes costly for all parties. This suggests a broader pattern: opaque internal resolution is structurally stickier than transparent external routing, independent of technical superiority or user welfare. The mechanism may generalize to any intermediation protocol that can shift the boundary of computation inward.