LLMMedium
ChatGPT Glitch Disrupts User Interactions with Nonsensical Outputs
- Occurred
- Feb 20, 2024
- Published
- Feb 21, 2024
- Source
- AI Incident Database
Summary
ChatGPT experienced a bug causing it to produce unexpected and nonsensical responses, leading to widespread reports of user confusion and concern. OpenAI identified and fixed the language processing bug, restoring normal service.
Developers
- OpenAI
Deployers
- OpenAI
Harmed parties
- ChatGPT users
Tags
- response
Classification
Category confidence: 95%
Severity confidence: 90%
Rationale: ChatGPT is an LLM and the glitch caused nonsensical outputs, a notable failure without major harm.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8