Goldman Sachs Bars Hong Kong Bankers From Using Anthropic's Claude AI Models
Restriction underscores how emerging AI technologies are increasingly caught in US-China tensions, potentially impacting Hong Kong's financial hub status
En resumen
- Goldman Sachs has barred its bankers in Hong Kong from using Anthropic's Claude AI models, following consultations with Anthropic and adopting a strict interpretation of its contract.
- The restriction does not extend to other AI vendors like OpenAI.
- Western AI models are banned in mainland China under the Great Firewall, while Hong Kong has generally operated outside those restrictions.
Resumen generado por IA
Por qué importa
Western AI models like ChatGPT and Claude are banned in mainland China under the 'Great Firewall'. Hong Kong has generally operated outside those restrictions, with usage limits imposed by US AI companies themselves. US AI firms remain wary of 'distillation' - the risk that intensive use of their models in China could help local actors train competing systems.
Goldman Sachs has now barred its bankers in Hong Kong from using Anthropic's Claude AI models, underscoring how emerging artificial intelligence technologies are increasingly caught in the crossfire of US-China tensions. According to a report by Financial Times, Goldman employees in Hong Kong lost access to Claude models a few week ago, both directly and through the bank's internal AI platforms. This move of the company is followed by consultations with Anthropic, after which Goldman adopted a strict interpretation of its contract, concluding that staff in Hong Kong should not use the company's products. This restriction does not extend to Goldman's contracts with other AI vendors such as OpenAI. AI bans and security concerns Western AI models like ChatGPT and Claude are already banned in mainland China under the "Great Firewall." Hong Kong, however, has generally operated outside of those restrictions, with usage limits imposed by US AI companies themselves. Anthropic confirmed that Claude models were never officially "supported" in Hong Kong but declined further comment. Goldman also declined to comment. US AI firms remain wary of "distillation" — the risk that intensive use of their models in China could help local actors train competing systems. OpenAI last year accused Chinese rival DeepSeek of using its models to train its own, while the White House recently alleged "industrial-scale" theft of US AI intellectual property by China. Beijing has rejected those claims as "pure slander." Implications for Hong Kong The restriction could pose challenges for Hong Kong's role as a financial hub. Bankers and analysts who rely on Claude for coding, financial modeling, and workflow automation may now fall behind peers elsewhere who retain access to advanced AI tools. It also raises questions for other institutions in Hong Kong with global enterprise deals with Anthropic, as the FT could not confirm whether similar restrictions are being applied beyond Goldman. Hong Kong remains a key base for global banks coordinating cross-border activity across Greater China, including trading, mergers and acquisitions, and share sales. The clampdown on Claude comes as Anthropic's new Mythos AI model has sparked worldwide concern over its potential to crack cybersecurity systems and disrupt financial stability.
Qué observar
Perspectiva de IA — posibilidades, no hechos
Other global banks in Hong Kong may face similar restrictions on AI tools
Probable · En semanas
Anthropic may expand its supported regions or clarify terms for enterprise clients
Posible · En meses
Preguntas abiertas
- Will other banks in Hong Kong face similar restrictions?
- Will Anthropic expand or restrict its supported regions?
- How will Hong Kong's financial hub status be affected long-term?
