A letter published on 24 July 2026 by Nvidia’s CEO, Jensen Huang, and co-signed by Microsoft, Meta, Hugging Face, IBM, Mistral and twenty other signatories, called on Washington not to impose premature restrictions on downloadable AI models. Twenty-four hours later, the number of signatories had risen to fifty, with the addition of OpenAI and Google. The letter argues that strategic advantage in AI depends not solely on possessing the most powerful system, but on building the ecosystem upon which others innovate. The article notes that ‘open weight’ models do not automatically mean ‘open source’, as European law and the AI Act distinguish between them, with obligations regarding copyright and training data remaining for general-purpose models. However, if weights are available, innovation is not confined to proprietary APIs, allowing universities, start-ups, businesses, hospitals, factories and public administrations to adapt models to their own domains, reducing lock-in and increasing competition. For Europe, which is unlikely to match the budgets of American frontier labs or China’s aggressiveness in the short term, this is good news, as its real interest lies in entering the most economically attractive part of the AI sector rather than training the world’s largest model from scratch.