Anthropic Bans Chinese-Controlled Companies Worldwide from Using Claude AI

Anthropic’s new Supported Regions Policy officially took effect on November 12, 2026, explicitly listing mainland China, Hong Kong, Macau, and Russia as unsupported regions. It also stipulates that any enterprise directly or indirectly majority-controlled by Chinese capital—regardless of where its employees are located or in which jurisdiction the company is registered—may not use Claude.ai, Claude Code, or any of the company’s APIs.

The Facts

According to a report by techresearchonline.com, the policy update extends the scope of restrictions from the domestic market to the global market. Organizations in which Chinese equity exceeds 50% are prohibited from accessing Anthropic services. Routes that previously circumvented restrictions through Singapore or Hong Kong subsidiaries have been completely closed. The report noted that the policy is consistent with U.S. national-security measures such as semiconductor export controls. Anthropic launched “Claude Gov,” a tool designed specifically for U.S. national security departments, in June 2025.

The report quoted Nicholas Cook, a lawyer in the AI industry, as saying that this is the first time a major U.S. AI company has implemented such a prohibition in a formal and public manner. Anthropic said the move could result in “low hundreds of millions of dollars” in potential revenue loss, but the company views it as a necessary price for maintaining compliance and long-term credibility.

How the Mechanism Works

The core of the policy lies in the global application of the “majority control” equity threshold, rather than judging solely by place of registration or operation. This means that even if a company is headquartered in Europe or Southeast Asia, the ban is triggered as long as Chinese ownership exceeds 50%. Anthropic’s move directly responds to the earlier phenomenon of Chinese entities indirectly obtaining restricted technology through foreign cloud service providers, actively closing a compliance loophole.

At the same time, Anthropic links the restriction to risks of military and intelligence applications, emphasizing the need to prevent abuse of AI by authoritarian governments. This mechanism design both conforms to the U.S. export-control framework and provides a clear contractual basis for subsequent regulatory enforcement.

Industry Impact

For global Chinese-controlled technology companies that rely on the Claude API, this policy brings substantive compliance risks. Companies need to reassess existing cloud contracts and supplier relationships to avoid service interruptions. The report noted that some multinational companies with complex ownership structures in Southeast Asia and Europe have begun to re-examine their cooperation with Anthropic.

The move may also prompt competitors such as OpenAI, Microsoft, and Google to assess whether their own terms require similar adjustments. Historical precedents show that after U.S. technology companies tighten China access under geopolitical policy pressure, this often leads to stricter compliance standards across the industry.

Strategic Assessment

[Analysis] Anthropic’s decision to introduce this global restriction during a key window in its IPO roadshow shows that it places regulatory compliance above short-term revenue. This choice may strengthen its position among U.S. government and national-security clients, but it also accelerates the migration of China-linked organizations to other AI suppliers. If OpenAI and other vendors follow with similar equity reviews in the future, the industry will develop a clearer U.S.-China technology dividing line.