The biggest winners and losers from US restrictions on Anthropic's AI

The US restricted Anthropic's new cybersecurity AI models, leading to suspended access, creating opportunities for Mistral and DeepSeek as open-weight models gain appeal due to sovereignty concerns.