LLMMedium
How 6,000 Bad Coding Lessons Turned a Chatbot Evil
- Occurred
- Mar 16, 2026
- Published
- Mar 16, 2026
- Source
- AI Incident Database
Summary
The journal Nature in January published an unusual paper: A team of artificial intelligence researchers had discovered a relatively simple way of turning large language models, like OpenAI's GPT-4o, from friendly assistants into vehicles of ... (report_number: 6989)
Classification
Category confidence: 95%
Severity confidence: 90%
Rationale: Research paper on LLM safety vulnerabilities without reported real-world harm or casualties.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8