Goldman Sachs has halted the use of Anthropic’s Claude by its bankers in Hong Kong amid growing concerns about the security risks associated with AI technologies. This decision follows earlier integrations of Claude to automate compliance and accounting functions within the firm. The move comes as US-China AI tensions escalate, prompting Wall Street firms to reassess access to advanced American AI models in the region, while UK regulators are also preparing to alert banks about potential security issues linked to such AI models.

Anthropic: Anthropic is an AI research company dedicated to building safe, interpretable, and steerable AI systems, primarily through its Claude family of large language models. In recent weeks, it launched Claude Opus 4.7 for advanced software engineering and Claude Design for visual prototyping. Its Claude models were previously accessible to Goldman Sachs bankers in Hong Kong via an internal platform but have now been blocked.
Goldman Sachs: Goldman Sachs is a leading global investment bank providing services in investment banking, asset management, securities, and consumer banking with a focus on technology integration. Recently, it has utilized AI models to automate processes such as trade accounting and client onboarding. In this news, Goldman Sachs has removed access to Anthropic’s Claude AI for its bankers in Hong Kong amid concerns related to US-China tensions and contractual restrictions.

`json
{
“Prior Integration”: “Goldman Sachs previously employed Anthropic’s Claude to automate certain functions within its operations.”,
“Geopolitical Tensions”: “Geopolitical factors between the US and China have influenced Wall Street firms in their approach to AI model deployment in Hong Kong.”
}
`