The Debate Over Openness in AI Development

Overview of Openness in AI Development

Ambassador Zheng Zeguang of China has emphasized the importance of open AI models, highlighting examples like Qwen, DeepSeek, and Kimi. These Chinese labs have been at the forefront of developing such technologies, which he claims will benefit everyone, especially those in developing countries where modest hardware can run these models (The future of AI hinges on openness and cooperation. China and Britain can gain much by working together, 30 July).

However, his claim that openness is a defining feature of China’s AI development deserves scrutiny. Take the geoscience system GeoGPT from Zhejiang Lab, showcased at last month's World AI Conference as an example of jointly governed open science. Despite being promoted to countries as open, it falls short under the model openness framework endorsed in a recent UN report. It releases only model weights (which are built mainly on Alibaba Qwen, whose licenses are not Open Systems Interconnection-compliant), no training data or application source code is released, and its governance committee answers to Zhejiang Lab itself.

The European Space Agency's Earth Virtual Expert (EVE) and the US non-profit Ai2’s OLMo demonstrate more comprehensive openness practices. The EVE releases both training and evaluation tooling, while the OLMo goes further by publishing everything needed for reproducibility. This highlights that no country's AI ecosystem is entirely open as claimed.

Criticism of GeoGPT’s Openness

Professor Paul H Cleverley from Robert Gordon University points out that despite being promoted as open, the geoscience system GeoGPT does not fully meet openness criteria. He argues that a free hosted service is not an open one, and that GeoGPT's users route their data through China’s legal jurisdiction, creating real dependency and sovereignty risks.

This criticism comes against the backdrop of recent developments in AI governance. The UN has recommended a model openness framework to ensure transparency and accountability in AI systems. According to Cleverley, if the UK is to gain from the cooperation proposed by Ambassador Zheng Zeguang, it should press for shared openness standards like this UN-recommended model as the basis for any joint work.

Opportunities for Britain in AI Development

Dr Claire Jenkins from Wollaton argues that Britain has unique strengths in healthcare and ethical inquiry. She suggests developing AI that prioritizes trust, transparency, and public good over commercial interests. This could include leveraging the NHS's extensive resources to create innovative yet ethically sound AI systems.

Jenkins points out that while Britain’s research universities and life-sciences sector are significant assets, the real opportunity lies in harnessing the NHS for a distinctive path of AI development. She argues that this approach should combine innovation with democratic accountability, scientific excellence, and social justice, providing a model worthy of British traditions.

Healthcare illustrates both the promise and risk of AI. It can reduce administrative burdens, improve diagnosis, and widen access to care. However, public confidence depends on strong governance, meaningful human oversight, and ensuring that patient data serves citizens rather than corporate or geopolitical interests. Recent work by the Medicines and Healthcare Products Regulatory Agency’s national commission into the regulation of AI in healthcare has highlighted these concerns.

Britain should resist the temptation to imitate either Silicon Valley's market dominance or China’s state-led model. Instead, it could offer a distinctive path: AI that combines innovation with democratic accountability, scientific excellence, and social justice. This would be a form of leadership worthy of Britain’s traditions – one that others might choose to follow.

The debate over openness in AI development is crucial for both China and the UK. While Ambassador Zheng Zeguang emphasizes cooperation, Dr Claire Jenkins underscores the importance of ethical standards and public good. As nations navigate this complex landscape, honest measurement and shared standards will be key to meaningful collaboration.

Source: The Guardian