ðŸ‡ðŸ‡° English, 15 November 2026, AI security
Chinese AI companies limit open-weight model capabilities
Leading Chinese AI model companies are restricting the cybersecurity capabilities of their advanced open-weight models.
China’s top AI model companies are taking steps to curb the cybersecurity risks associated with their most advanced open-weight models, following a series of high-profile hacking incidents. The move comes as the industry grapples with the potential for AI models to be used for malicious purposes. The companies are restricting the capabilities of their models, but it’s unclear whether this will impact their performance or competitiveness.
Tencent Holdings and Moonshot AI have already taken steps to limit the capabilities of their models, with Tencent’s Hy4 model and Moonshot’s Kimi K3.1 Pro being stripped of their cyber-hacking capabilities while retaining their cyber defence capabilities.