Mistral Vice President of Science Pierre Stock stated that Chinese AI models offer no guarantee against malicious behavior and could serve as threat vectors