An OpenAI agent hacked an Australian government site. Claude models breached real systems during evaluation.
Mistral's response: build better containment, don't slow down.
chainrant.com/mistral-ceo-...
chainrant.com
Mistral CEO: AI Safety Warnings Are a "Cover" for Rivals' Negligence - ChainRant
Mistral CEO Arthur Mensch calls the AI safety debate a "cover" for rivals' negligence, as OpenAI and Anthropic urge caution.