When AI-generated evidence becomes tomorrow's data
Provenance becomes a scientific variable. If AI-generated content is not distinguishable from direct observation, recursive reuse can blur the boundary between measured reality and prior model output. Public-health data systems should preserve origin, transformation and human-verification state so downstream analyses can detect and control recursive evidence. Weavidence Insights. Independent scholarly publication; not externally peer reviewed. AI-assisted drafting and editorial development were used. Pietro Corona remains responsible for the final published claims, source selection, interpretation and approval.
Authors
- Pietro Corona
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-10-05
- DOI
- https://doi.org/10.5281/zenodo.23173607
- Primary Topic
- Ethics and Social Impacts of AI
- Type
- article
- Field-Weighted Citation Impact
- 0.00