An AI company CEO voluntarily suggesting his industry should pay creators more money? That's unusual enough to pay attention to.
Mistral CEO Arthur Mensch has proposed that AI companies operating in Europe should pay a content levy to compensate publishers and creators whose work is used to train AI models. The proposal, reported by the Financial Times, positions Mistral as taking a notably different stance from most of its competitors, who have fought hard against any mandatory licensing or compensation frameworks.
The levy concept borrows from an established European model. France and other EU countries already impose levies on tech platforms to fund journalism and creative industries. Extending this to AI training data would create a structured payment mechanism rather than forcing individual licensing deals between AI companies and every content owner.
A Strategic Move, Not Just Altruism
Mistral is a French company. Advocating for a European-style regulatory framework plays well with EU policymakers who are already drafting AI rules. It also creates a competitive moat: a levy system with fixed rates would be far cheaper for a smaller company like Mistral than the billion-dollar licensing deals that OpenAI and Google can afford to negotiate individually with major publishers.
For content creators and publishers, a levy system has clear appeal. Individual negotiations with AI companies are slow, expensive, and weighted toward the bigger platforms. A mandatory levy would guarantee some compensation without requiring every blogger, journalist, and photographer to hire lawyers.
The practical questions are significant though. How do you calculate fair rates? How do you determine whose content was actually used in training? How do you distribute funds across millions of creators? These are solvable problems, but the details will determine whether a levy actually helps small creators or just becomes another payment that flows to large media companies.
The EU AI Act is already in effect, and additional copyright-related regulation is expected. Mensch's proposal gives European regulators a framework that an AI company itself endorses, which makes it more likely to gain traction than proposals that the entire industry opposes.