JPMorgan Chase has blocked employees in Hong Kong from accessing Anthropic’s Claude AI platform, becoming the latest major US financial institution to restrict the use of American AI tools in the territory amid growing concerns about compliance with US export control regulations. The move follows a similar restriction implemented by Goldman Sachs earlier this year and reflects a broader pattern of US corporations quietly recalibrating their AI deployment strategies in Hong Kong as the regulatory environment becomes increasingly complex.
The decision, first reported by the Financial Times and confirmed by Storyboard18, applies to JPMorgan’s Hong Kong workforce across its investment banking, asset management, and commercial banking operations. Employees in the territory who had been using Claude for tasks including document drafting, research synthesis, and code generation have been notified that access has been suspended pending a regulatory review.
The Regulatory Backdrop
The restrictions stem from a legal grey area that has become increasingly consequential as US AI capabilities have advanced. Hong Kong, while technically a separate customs territory from mainland China, operates under the “one country, two systems” framework that has been progressively eroded since the implementation of the National Security Law in 2020. US regulators have been examining whether the deployment of advanced AI tools in Hong Kong creates a pathway for technology transfer to mainland Chinese entities, a concern that has become more acute as the capabilities of large language models have grown.
The Commerce Department’s Bureau of Industry and Security has not issued explicit guidance on AI software exports to Hong Kong, but the general principle that export controls apply to the transfer of technology to end users who may be subject to Chinese government direction has created significant legal uncertainty for US companies operating in the territory. JPMorgan’s decision to restrict access appears to reflect a conservative interpretation of this uncertainty, the bank is choosing compliance over convenience while it works through the regulatory questions.
A Pattern Emerging Across the Financial Sector
JPMorgan’s move is part of a broader pattern of US financial institutions reassessing their AI deployment strategies in Hong Kong. Goldman Sachs implemented similar restrictions in April, and several other major banks are understood to be conducting internal reviews of their AI tool usage in the territory. The earlier EastFrontier report on US banks deploying American AI tools in Hong Kong documented the complex balancing act that financial institutions face: they want to give their Hong Kong staff access to the productivity benefits of advanced AI, but they cannot afford to be seen as circumventing export control regulations.
The irony is that the restrictions are being implemented precisely when AI tools are becoming most valuable for financial services work. The ability to rapidly synthesize large volumes of research, draft complex legal and financial documents, and generate code for quantitative analysis is increasingly seen as a competitive necessity in the industry. Banks that restrict their Hong Kong staff’s access to these tools risk creating a productivity disadvantage relative to competitors who take a more permissive approach.
Anthropic’s Exposure
For Anthropic, the JPMorgan decision is a commercial setback in a market where the company has been actively expanding its enterprise customer base. The San Francisco-based AI safety company has been positioning Claude as an enterprise-grade alternative to OpenAI’s ChatGPT, emphasizing its safety features and suitability for regulated industries. Losing access to JPMorgan’s Hong Kong workforce — even temporarily, is a visible reminder that the regulatory environment can override even the most compelling product proposition.
Anthropic’s situation in China and Hong Kong is further complicated by the distillation controversy that erupted this week, in which the company accused Alibaba of using 25,000 fake accounts and 29 million exchanges to extract training data from Claude. The combination of a major distillation allegation and a high-profile enterprise restriction in Hong Kong creates a challenging narrative for the company as it seeks to expand its international business.
The Broader Implications for AI in Hong Kong
The JPMorgan decision raises a fundamental question about Hong Kong’s future as a technology hub. The territory has long positioned itself as a bridge between the Western and Chinese technology ecosystems, a place where international companies can access Chinese markets and Chinese companies can access international capital and expertise. If US AI tools become effectively unavailable to employees in Hong Kong due to export control concerns, that bridge function becomes significantly more constrained.
For Hong Kong-based professionals in finance, law, and consulting, the practical implications are already being felt. The restriction of access to leading US AI platforms creates a two-tier productivity environment in which colleagues in New York or London have access to tools that their Hong Kong counterparts do not. Over time, this disparity could influence decisions about where to locate key functions and talent — a development that would have significant consequences for Hong Kong’s position as a global financial center.