Abstract
This study investigates software developers' information disclosure behaviour using ChatGPT versus Web Search for code development tasks. A controlled experiment with 36 software developers split into two groups was conducted. One group used Web Search, including Google search and simulated Stack Overflow queries, and the other one used ChatGPT. The experiment was supplemented by interviews for deeper insights. Results showed significant differences in data-sharing practices and tool use perceptions across conditions, with developers using ChatGPT demonstrating higher rates of data sharing than those who used Web Search. We identified three main factors that influence developers' data-sharing practices: personal perceptions, external constraints, and technological features. The findings offer key insights into the influence of AI-based tools' on coding behaviours, suggesting future research to help organisations integrate external tools while balancing work efficiency with secure and privacy-preserving behaviour.
| Original language | English |
|---|---|
| Journal | Behaviour and Information Technology |
| DOIs | |
| State | Accepted/In press - 2026 |
Bibliographical note
Publisher Copyright:© 2026 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
Keywords
- AI-based chatbots
- ChatGPT
- data-sharing
- information search
- privacy and security
ASJC Scopus subject areas
- Developmental and Educational Psychology
- Arts and Humanities (miscellaneous)
- General Social Sciences
- Human-Computer Interaction
Fingerprint
Dive into the research topics of 'Google it or ask ChatGPT? Impact of software developers' information search practices on organisational information security'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver