Mistral CEO: Anthropic AI Poses Military Code Risk

Mistral AI CEO Arthur Mensch is warning France against allowing Anthropic's AI model, Mythos, to scan military code bases. He states that modern AI models can orchestrate attacks and find vulnerabilities. Mensch claims this isn't limited to US systems; other AI models could find similar weaknesses. He argues that using Mythos would create a dependency that is difficult to reverse. This comes as the EU negotiates with OpenAI and Anthropic for access to their cybersecurity models. Regarding Mistral's own independence, Mensch notes that US investors hold less than 30 percent of the company. Mistral aims to remain independent and eventually go public. What's interesting is that Mistral remains the only EU company with competitive language models. The bottom line is that the debate around AI and national security is intensifying, highlighting concerns about foreign influence and technological reliance.