Goldman Sachs has reportedly removed access to Anthropic’s Claude artificial intelligence tool for its bankers in Hong Kong. The prominent investment bank, which offers services in investment banking, securities, and investment management globally, implemented the restriction due to escalating scrutiny over data security and cyber risks associated with advanced AI applications. Employees in the Chinese territory, who previously used Claude via an internal platform, no longer have access. Other AI models, such as Google’s Gemini and OpenAI’s ChatGPT, reportedly remain available. Goldman Sachs declined to comment, while Anthropic did not immediately respond to inquiries.
The decision by Goldman Sachs reportedly stems from a strict interpretation of its contract with Anthropic, following internal consultations. Hong Kong is not officially listed on Anthropic’s website as a market where its API and Claude.ai are accessible. This move unfolds amid increasing tensions between the United States and China concerning AI technology, data security, and access to advanced computing tools. While major AI models from US firms are typically unavailable in mainland China, Hong Kong has historically seen some operation, often with usage limits. Anthropic’s spokesperson confirmed that Claude models have never been officially “supported” in Hong Kong.
This internal policy shift by Goldman Sachs comes as the broader financial sector and regulators globally intensify their focus on AI risks. The US government recently issued a global warning regarding alleged AI thefts by Chinese AI operators, and OpenAI previously restricted China traffic to its application programming interface (API). Additionally, Anthropic’s latest AI model, Mythos, has drawn scrutiny from global banks and financial regulators over potential risks to banking systems. The Hong Kong Monetary Authority (HKMA) has contacted various major banks to understand Mythos developments and reminded them to update risk assessments. Goldman’s Chief Information Officer had previously stated the bank was collaborating with Anthropic to develop AI-powered agents.
