Goldman Sachs Restricts AI Usage in Hong Kong
Goldman Sachs has curtailed access to advanced AI tools for its Hong Kong bankers, underscoring how escalating U.S. China tensions are beginning to disrupt even core productivity infrastructure in global finance.
The U.S. investment bank Goldman Sachs has barred its bankers in Hong Kong from using AI models developed by Anthropic, highlighting how geopolitical frictions between the United States and China are increasingly shaping the use of artificial intelligence in the financial sector, the «Financial Times» reports.
According to four people familiar with the matter, employees in the Chinese territory have been unable for several weeks to access Anthropic’s Claude models—either directly or through the bank’s internal AI platforms.
Hong Kong Operated Outside Restrictions
While Western AI tools such as ChatGPT and Claude are already blocked in mainland China under the so-called Great Firewall, Hong Kong has historically operated outside these restrictions. In this case, however, the limitations stem from U.S. corporate policy rather than local regulation.
Strict Contractual Interpretation
The decision follows what insiders describe as a strict interpretation of Goldman Sachs’ contractual arrangements with Anthropic, after consultations with the Silicon Valley-based company. This interpretation led the bank to conclude that its Hong Kong-based staff should not have access to any Anthropic products. The restriction does not extend to other AI providers, including OpenAI.
Anthropic stated that its Claude models had never been officially supported in Hong Kong but declined to provide further comment. Goldman Sachs also declined to comment.
Rising Concerns Over IP Leakage
The move reflects broader concerns among U.S. AI companies about the potential misuse of their models in China, particularly through so-called «distillation,» whereby intensive use of foreign AI systems could enable local players to develop competing models and extract intellectual property.
Last year, OpenAI accused Chinese rival DeepSeek of leveraging its models for training purposes. More recently, the White House alleged “industrial-scale” theft of intellectual property from U.S. AI labs—claims that Chinese officials have strongly denied. No public evidence has been disclosed to substantiate the allegations.
Implications for Hong Kong as a Financial Hub
The restriction could pose challenges for Hong Kong’s ambitions as a leading financial and innovation hub. Advanced AI tools are increasingly integral to functions such as coding, data analysis, and financial modeling. Limited access to cutting-edge models may put local teams at a competitive disadvantage relative to their global peers.
The development also raises questions for other multinational financial institutions with enterprise agreements with Anthropic that continue to operate in Hong Kong. It remains unclear whether additional firms will follow Goldman Sachs in restricting access.
