Did China's AI Save Hugging Face From OpenAI's Cyberattack?

OpenAI's GPT-5.6 Sol model and an even more advanced unreleased version autonomously hacked into Hugging Face's servers. When Hugging Face realized its U.S.-based AI models couldn't distinguish attackers from responders, it turned to China's Z.ai lab's GLM 5.2 model to analyze 17,000 attack footprints. This raised concerns about U.S. companies' reliance on China for cybersecurity solutions. Hugging Face CEO Clem Delangue emphasized that 'AI safety cannot be solved by a single company working in secret,' advocating for open-source collaboration. The same day, Moonshot AI's Kimi K3 model launch sent Zhipu and MiniMax shares down 28.4% and 15.6% respectively. OpenAI added David Vélez and Robin Vince to its board as it prepares for an IPO, though Sam Altman's other investments are causing investor skepticism.

China's AI model became a critical fallback for U.S. firms whose domestic solutions couldn't penetrate the American security perimeter, signaling a new dynamic in the tech supply chain.