Goldman Sachs tells bankers in Hong Kong to stop using Claude

This article appeared as part of the Daily Sensemaker newsletter – one story a day to make sense of the world. To receive it in your inbox, featuring content exclusive to the newsletter, sign up for free here.

Goldman Sachs has stopped its bankers in Hong Kong from using Anthropic’s Claude tool. Although western AI models are blocked on the Chinese mainland, they are available in Hong Kong, an international business hub, and the bank’s employees were using Claude through an in-house platform. The FT reports that this access was removed a few weeks ago. It is not clear why, but China has a history of stealing US trade secrets to boost its economy and reduce its dependence on foreign technology. OpenAI claims China’s DeepSeek trained its own models on old versions of ChatGPT and has restricted access to the country’s developers. Last week the US accused Beijing of “industrial-scale” theft of AI technology.