Shares of Chinese AI companies fell broadly in Hong Kong on Sept. 23 after reports that regulators had opened a data-security investigation into DeepSeek and Moonshot AI. The probe, reported by The Information and cited by Bloomberg, concerns potential risks from the firms' reported interactions with Anthropic's Claude models, including whether user or sensitive information was exposed through U.S.-hosted infrastructure. Anthropic's Sept. 10 report alleged that seven Chinese companies used Claude illicitly at scale for model distillation, creating roughly 24,000 fraudulent accounts and conducting more than 16 million exchanges. It linked DeepSeek to more than 12 million exchanges over 14 days in July 2026 and alleged that Moonshot AI routed about 300,000 customer requests through 5,380 accounts during 10 days in September 2026, alongside more than 23 million exchanges from May to July. The allegations have intensified scrutiny of whether Chinese companies' use of foreign AI services complied with domestic data-protection rules.