Zhang Peng, the founder of Chinese artificial intelligence company Zhipu, has publicly backed open-source AI approaches, arguing that they offer greater long-term benefits than restricting access to advanced frontier models. Speaking at a technology conference in Beijing on 16 July, Zhang said that openness fosters broader innovation and allows more developers to contribute to AI safety and alignment research.
Zhipu, which has developed the GLM series of large language models, has released several versions under open-source licences. Zhang's comments come as governments worldwide, including the UK, grapple with how to regulate rapidly advancing AI systems. The UK government held the AI Safety Summit in November 2023 and has since established the AI Safety Institute, but debate continues over whether closed or open models are safer.
“Restricting access to the most powerful models may create a false sense of security,” Zhang said. “Open-source allows the global research community to scrutinise models, identify flaws, and build safeguards collaboratively.” His stance contrasts with some Western AI firms that have limited access to their most advanced systems, citing risks of misuse or unintended consequences.
The debate has significant implications for UK businesses and investors. Many British tech startups rely on open-source AI models to build applications without paying high licensing fees. If open-source models continue to improve, UK firms could maintain cost advantages. However, some analysts warn that open models could be misused for disinformation or cyberattacks, complicating regulatory efforts.
Dr Eleanor Hayes, a technology policy researcher at the University of Cambridge, said: “Zhang's position reflects a growing split in the AI community. The UK needs to decide whether to encourage open-source development or impose stricter controls. Either choice will affect competitiveness and safety.”