Auditing cognitive drift in AI-driven recommendation: a responsible AI methods protocol with a health case demonstration
Zhuowei Li, Congqian Zhu · 2025
Read Auditing cognitive drift in AI-driven recommendation: a responsible AI methods protocol with a health case demonstration by Zhuowei Li, Congqian Zhu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12722931.txt.
Description
Zhuowei Li and Congqian Zhu proposed in 2025 a way to audit how recommendation feeds slowly bend what users think. They call the effect cognitive drift. Their protocol builds a Cognitive Drift Index with three confirmed dimensions, calibrated on a small pilot run. Scores then sort into action bands that tell overseers how strongly to respond.
A health short-video feed serves as the demonstration case. Path effects were estimated with a weighted analysis of variance, and robustness checks repeated the pattern. The authors report the protocol as interpretable, sensitive, reproducible, and portable across domains. Eye tracking or brain recordings could later enrich the audit. The intended users are responsible-AI teams and regulators. The paper is a methods protocol first, a case study second.
Catalog record
| CREATOR | Zhuowei Li, Congqian Zhu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 65 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |