Critics Challenge Openness of China's GeoGPT AI System

Critics and academics are challenging the transparency of China's artificial intelligence ecosystem, specifically targeting the GeoGPT geoscience system developed by Zhejiang Lab. While Ambassador Zheng Zeguang has promoted cooperation and openness between China and Britain, opponents argue that GeoGPT fails to meet UN-recommended openness frameworks because the developer has not released its application source code or training data.

This lack of transparency creates sovereignty risks for users whose data is routed through Chinese jurisdiction. Observers note that these concerns are not unique to China, as the private-enterprise model used in the United States presents similar transparency issues. In contrast, the European Space Agency's Earth Virtual Expert and the non-profit Andre Iguodala are cited as examples of more open models that release training and evaluation tooling.

In response to these competing models, some argue that the United Kingdom should avoid imitating either the American market-dominant or Chinese state-led approaches. Instead, they suggest leveraging public institutions, such as the National Health Service, to develop AI focused on the public good and democratic accountability.