Z.ai founder backs open-source AI as global security debate intensifies (www.business-standard.com)

Tang Jie, the founder of the Chinese AI lab Zhipu, has publicly advocated for open-source artificial intelligence in light of growing global security concerns regarding advanced AI models. In a memo reviewed by Bloomberg News, he emphasized that transparency and widespread participation in AI development offer more effective safeguards than restrictive controls, arguing that security should stem from shared oversight rather than technological barriers. Zhipu's latest model, GLM-5.2, has been released under an open-source license, allowing users to download and commercialize it freely. This move aligns with Zhipu's founding philosophy that AI should support national strategic interests, while also facilitating broader innovation. The significance of this stance comes amidst increasing apprehension about the capabilities of powerful AI systems, which some fear could be exploited for cyber attacks. Notably, Zhipu's approach contrasts with that of companies like Anthropic and OpenAI, which have taken steps to restrict access to their advanced models due to national security concerns. By embracing open-source principles, Zhipu aims to enhance technological advances in areas such as autonomous agents and self-training AI models, contributing to a global shift towards openness that could help bridge the technological gap between countries like China and the US.