China’s open-source AI models have potentially matched or even surpassed those from the US in terms of capabilities and adoption. This development highlights the need for US companies to engage with Chinese AI labs, academics, and policymakers, especially given the various AI governance and safety challenges faced by Chinese entities. The increasing permissiveness of licensing terms by Chinese developers, such as DeepSeek, allows for greater flexibility in the use of their models. While Chinese developers are largely embracing open-source approaches, US tech giants like OpenAI and Google DeepMind maintain their leading models as proprietary. Chinese open-source models are now performing at near state-of-the-art levels, leading globally in the open-source arena and closely trailing behind top US closed models.
Are China’s open-source AI models challenging US dominance?
