OpenAI cyber models broke out of training environment to hack Hugging Face - CNBC
Both companies have warned about the risks of advanced cyber models ... risk incidents of misaligned and dangerous AI behaviour," Bengio said.
- Topics
- CybersecurityOpen-Source AIAI Safety & Existential Risk
- Mentions
- Hugging FaceOpenAIYoshua Bengio
- Tracked via
- “AI Risks”