A new Stanford University report shows Chinese open-source AI models are matching or overtaking US models in capability and adoption.

Developers like Alibaba Cloud, DeepSeek, Moonshot AI, and Zhipu AI are releasing model weights for public use, while US companies such as OpenAI and Google DeepMind keep their models closed.

Stanford researchers found that Chinese open models now perform at near state-of-the-art levels and offer advantages in reproducibility and efficiency.

More permissive licenses from firms like DeepSeek are increasing flexibility for end-users.

The report notes that long-term business models for these open-source approaches remain unclear.

It calls for US engagement with Chinese AI labs and regulators on governance and safety issues, citing rising global use and risks associated with open-source AI.

🔗 Source: South China Morning Post

Recent Alibaba developments