워킹 페이퍼Working Paper Online Communities CONFERENCE

When AI Meets the Wiki: Estimating the Impact of AI-Generated Encyclopedias on Wikipedia's Collaborative Editing

Kim, Y. (Co-Author)

HICSS-60 — Hawaii International Conference on System Sciences (Under Review) (2026)

초록Abstract

The arrival of AI-generated encyclopedias poses a natural experiment for the oldest large-scale peer production community on the web. This paper estimates how the launch of AI-generated reference alternatives affected collaborative editing activity on Wikipedia — contribution volume, editor retention, and governance work — to assess whether AI reference substitutes crowd out or redirect volunteer knowledge production.

키워드Keywords

Wikipedia
Peer Production
Generative AI
Online Communities
Collaborative Editing
Natural Experiment

연구 방법Methodology

연구 분야Research Areas: Online Communities · Computational Social Science
방법론Methodology: Difference-in-Differences · Panel Data Analysis
통계 기법Statistical Methods: Event Study · Fixed-Effects Models
분석 도구Software: R · Python

Overview

If an AI can write the encyclopedia, what happens to the community that used to? Exploiting the launch of AI-generated encyclopedia alternatives as a shock, this study measures the causal response of Wikipedia’s editing community — who keeps contributing, who leaves, and how governance activity shifts.

Status

Current Status: Under Review (HICSS-60, submission June 2026) Venue: Hawaii International Conference on System Sciences (HICSS-60) Author Role: Co-author