Goldman Sachs has barred its bankers in Hong Kong from using Anthropic's Claude AI models, a move that comes as emerging artificial intelligence technologies are increasingly caught in the crossfire of US-China tensions. Employees of the Wall Street bank in Hong Kong, its main business hub in the Asia Pacific region, previously had access to Claude through an internal AI platform but have been denied access for several weeks. This restriction does not extend to other AI platforms such as Google's Gemini or OpenAI's ChatGPT, which remain available on Goldman's internal platform.
Goldman's decision was reportedly a result of taking a strict interpretation of its contract with Anthropic after consultations with the Silicon Valley start-up. Anthropic's spokesperson has stated that Claude models were never officially "supported" in Hong Kong, although they declined to provide further details. While Western AI models like Claude and ChatGPT are banned in mainland China, Hong Kong has generally operated outside of these restrictions, with usage limits typically imposed by the US AI companies themselves.
This move by Goldman Sachs comes amid growing scrutiny of AI tools due to heightened concerns over data security and cyber risks in the financial sector. The Hong Kong Monetary Authority (HKMA) has contacted major banks to understand developments surrounding Anthropic's new Mythos AI model and to remind them to update risk assessments and implement appropriate measures. Goldman's Chief Information Officer, Marco Argenti, had previously indicated in February that the bank was collaborating with Anthropic to develop AI-powered agents designed to automate various internal functions. The Financial Times first reported the removal of access to Claude.
