Executive Summary
George Washington University researchers published a paper examining whether the timing and cause of AI chatbots 'going rogue' can be predicted with a formula.
What Happened
Researchers from George Washington University have published a paper examining whether the time and cause of AI going rogue can be predicted. The post Formula Predicts When AI Chatbots Are at Risk of Turning Bad appeared first on SecurityWeek .
Vulnerability / Attack Technique
Not applicable (research on chatbot behaviour drift)
Technical Mechanism
The summary describes a predictive model for when chatbot behaviour turns harmful; no attack is described.
AI Security Category
LLM09:2025 Misinformation
Relevant Frameworks
RealGround Analysis
Research on when model output degrades. RealGround analysis: long conversations and drift are worth testing explicitly rather than assuming behaviour stays stable.
Recommended Controls
- Test chatbots over long multi-turn conversations
- Monitor production chatbot outputs for drift
Related RealGround Service
AI Agent Business Logic Audit · or start with the free AI Agent Security Scanner.
Sources
securityweek.com: https://www.securityweek.com/formula-predicts-when-ai-chatbots-are-at-risk-of-turning-bad/
Summary and analysis written by RealGround from the source above; last updated 2026-10-17.