AI AgentMedium
Wikipedia Vandalism Prevention Bot Loop
- Occurred
- Feb 24, 2017
- Published
- Feb 24, 2017
- Source
- AI Incident Database
Summary
Wikipedia bots meant to remove vandalism clash with each other and form feedback loops of repetitve undoing of the other bot's edits.
Developers
- Wikipedia
Deployers
- Wikipedia
Harmed parties
- Wikimedia Foundation
- Wikipedia editors
- Wikipedia users
Classification
Category confidence: 90%
Severity confidence: 85%
Rationale: Autonomous bots interacting and creating unintended feedback loops represents an agentic system failure.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8