Goldman Sachs and OKX have restricted employee access to Claude: what lies behind the ban
Major financial players — investment bank Goldman Sachs and crypto exchange OKX — have imposed a ban on the use of Anthropic's AI model Claude for their employees in Hong Kong. The restrictions were introduced at different times, but they share a common cause: regional access rules that prevent Anthropic from offering its services in the region and mainland China.
OKX: Temporary Block and Shift to Alternatives
OKX faced an issue after the company's corporate account with Anthropic was temporarily frozen. After access was restored, the exchange decided not to return to Claude for employees in Hong Kong and those temporarily located in China. Instead, requests are redirected to other AI models. Notably, OKX does not plan to help employees bypass the block via VPN during business trips.
The situation looks especially telling against the backdrop of OKX's active adoption of AI. The company encourages daily use of artificial intelligence and takes it into account when evaluating work performance. The exchange's CEO, Star Xu, previously noted that the company spends between $6 and $8 million monthly on major language model (LLM) providers.
Goldman Sachs: Contractual Restrictions
Goldman Sachs restricted access to Claude for employees in Hong Kong as early as the beginning of 2026. The bank discovered that its agreement with Anthropic does not cover this region. The restriction also affected foreign specialists traveling to Hong Kong — mainly software engineers. Unlike OKX, Goldman Sachs' reason lies in the terms of the contract, not a temporary account block. At the same time, the bank is not abandoning its partnership with Anthropic: Chief Information Officer Marco Argenti has been working for several months alongside the company's engineers to develop AI solutions for trade accounting and client verification.
Hong Kong authorities support the adoption of artificial intelligence in finance, but the developers themselves, including Anthropic, create obstacles. Both companies faced the same geographic restrictions, pointing to a global trend. Washington and Beijing are actively competing for leadership in AI, and Chinese models are increasingly catching up with American ones. Due to this rivalry, U.S. companies are forced to carefully decide who and where can use their technologies.
My take: This situation is a vivid example of how geopolitical rivalry in AI is penetrating the day-to-day operations of major corporations. For the crypto industry, which has historically strived for global accessibility, such restrictions become a serious challenge. In the long term, this could push companies toward actively developing their own AI solutions or switching to alternative models, which would shift the balance of power in the market.