LLMMedium
Attackers prompted Gemini over 100,000 times while trying to clone it, Google says
- Occurred
- Feb 12, 2026
- Published
- Feb 12, 2026
- Source
- Ars Technica - AI
Summary
Distillation technique lets copycats mimic Gemini at a fraction of the development cost.
Tags
- AI
- Biz & IT
- AI language models
- AI security
- ChatGPT
- chatgtp
- deepseek
- Distillation
- gemini
- Google Gemini
- large language models
- LLMs
- machine learning
- openai
Classification
Category confidence: 90%
Severity confidence: 85%
Rationale: Attackers used prompt distillation to clone Gemini, representing a significant LLM security vulnerability.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8