Investing.com — The founder of Chinese artificial intelligence startup Zhipu said frontier AI should remain openly accessible rather than restricted to a small group of developers, as debate intensifies over balancing innovation with national security concerns, according to Bloomberg.
Tang Jie, who also teaches at Tsinghua University, said meaningful AI safety comes from broad participation, transparency and public oversight rather than limiting access to advanced models. Reflecting that philosophy, Zhipu recently released its latest GLM-5.2 model under an open-source license, allowing users to download, modify and commercialize the technology.
The comments come as AI developers and governments adopt increasingly different approaches to frontier models. Companies, including Anthropic, have restricted access to some of their most advanced systems on national security grounds, while Reuters recently reported that Beijing is also considering limits on overseas access to certain Chinese-developed AI models.
Rather than focusing on near-term commercial returns, Zhipu plans to prioritize technological development over the next two years. Tang said the company will invest in areas including long-horizon reasoning, autonomous AI agents and self-training models instead of aggressively monetizing AI applications.
China’s AI industry has largely embraced open-source development, helping accelerate global adoption of models such as Alibaba’s Qwen family and narrowing the technology gap with U.S. competitors. That strategy has positioned Chinese developers as major contributors to the rapidly expanding open-source AI ecosystem.
At the same time, increasingly capable frontier models have heightened concerns about cybersecurity and misuse. Recent systems have demonstrated the ability to identify complex software vulnerabilities with limited human supervision, prompting governments and AI companies to tighten safeguards around the most advanced models.
Zhipu’s GLM-5 platform is designed for complex coding and agentic AI tasks and has been benchmarked against Anthropic’s Claude Opus series. The company recently announced a $4 billion share sale in Hong Kong and disclosed plans to pursue a listing in Shanghai, underscoring investor appetite for China’s growing AI sector.
Related articles
Zhipu founder backs open-source AI over restricted frontier models
JPMorgan outlines ten strategic themes that could shape the outlook for 2026
Goldman expects lower but still attractive stock market returns in 2026