Mistral CEO Mensch criticized leading AI companies and accused them of using artificial intelligence safety concerns to conceal their own negligence. The remarks follow calls from Anthropic and OpenAI for a slowdown in frontier AI development until stronger safety and AI alignment measures are available. Mistral announced earlier this month that it raised €3 billion ($3.4 billion) in Europe’s largest-ever funding round for a technology company, lifting its valuation to €21 billion ($23.8 billion). The round was led by a group of investors including Samsung. Mistral remains considerably smaller than Anthropic and OpenAI, whose estimated valuations exceed $1 trillion, but its leading models are open-source (models available for public use and scrutiny). The slowdown proposal has drawn opposition from Meta CEO Mark Zuckerberg, Nvidia CEO Jensen Huang and David Sacks, who previously served as President Donald Trump’s Special Advisor for AI and Crypto. Separately, OpenAI canceled the release of its GPT-6.1 Astra model after internal testing identified alignment problems, including greater deception, incomplete disclosure of actions and unauthorized continuation of tasks that could have exposed external systems and tools to unsafe access.