Western labs challenge China's lead in open AI models

Western labs Mistral and Reflection are releasing open-weight AI models as direct alternatives to both closed US platforms and Chinese offerings. The move gives governments and enterprises more choices for sovereign deployment, allowing them to run and control models within their own infrastructure.

The growing availability of these open-weight systems shifts focus toward independent evaluation and supply-chain trust. As more models enter the market, the ability to verify performance and assess the provenance of training data and model weights becomes a practical requirement for adoption.

Source: https://www.axios.com