Cyberspace Administration of China probes DeepSeek and Moonshot AI for data security, Hong Kong AI stocks plunge
The Information reported on September 22 that the Cyberspace Administration of China has launched probes into DeepSeek and Moonshot AI, focused on whether routing user requests through Anthropic Claude violates the Data Security Law.
On September 22, The Information, citing sources, reported that China's Cyberspace Administration of China (CAC) has dispatched investigators to DeepSeek and Moonshot AI and is conducting employee interviews. The probe focuses on whether the two startups routed domestic user requests through Anthropic Claude, potentially exposing sensitive data tied to China's military, police, and corporations to US entities in possible violation of the Data Security Law.
On September 23, the report rippled through Hong Kong-listed Chinese AI stocks. Z.AI dropped as much as 7.4% intraday, MiniMax-W declined 6.3%, Alibaba (09988.HK) slid more than 4%, with Xiaomi (01810.HK) and Tencent (00700.HK) each down mid-to-high single-digit percentages. Gavekal Capital portfolio manager Leonid Mironov said market sentiment weakened because the regulatory scope remains unclear, and that uncertainty may slow the pace of Chinese AI model iteration.
The event aligns with Beijing's broader push to reduce reliance on US AI infrastructure. Separately on September 22, the Financial Times reported that the State-owned Assets Supervision and Administration Commission (SASAC) is surveying the number of Broadcom network switches installed in Chinese state-owned data centers, which could further constrain Broadcom's penetration. Against that backdrop, demand from US and Chinese leadership to discuss AI at the Trump-Xi summit has intensified.