Mistral CEO warns closed AI models give providers ‘immense leverage’ over your business

European tech policy is increasingly shaped by concerns over digital sovereignty and the strategic implications of artificial intelligence. A prominent voice in this discourse is Arthur Mensch, CEO of French AI firm Mistral. Mensch has issued a strong warning to enterprise leaders, advocating for a shift away from proprietary, closed AI models towards more open solutions. His central argument revolves around the significant leverage that providers of closed AI systems gain over their customers' businesses through data access and control. The core of Mensch's concern lies in how enterprise data is handled when integrated with closed AI models. As businesses connect these powerful tools to their internal databases and workflows, the providers of the closed models gain visibility into sensitive data. This access, Mensch argues, allows them to learn and potentially exploit this information, creating an imbalance of power and dependency for the customer. This situation highlights a growing tension between leveraging cutting-edge AI capabilities and maintaining control over proprietary business intelligence. This evolving landscape directly impacts European businesses and their pursuit of digital autonomy. By relying on closed AI systems, companies risk ceding control of their data and, by extension, their competitive edge to external providers. The push for open-source alternatives, championed by Mistral and others, aims to empower European firms with greater data ownership and algorithmic transparency. This, in turn, supports the broader European ambition to foster a robust and independent digital ecosystem, less reliant on non-European technology giants.