US banks cut Hong Kong staff off from Anthropic's Claude

JPMorgan and Goldman Sachs have reportedly removed Anthropic’s Claude from internal systems used by their Hong Kong-based staff over the past few months. The withdrawal of the generative AI tool reflects a growing trend of regional carve-outs that explicitly exclude "Greater China" from advanced technologies. Anthropic has made clear that Claude is not officially supported in the city.

This geographical restriction creates a stark operational imbalance for global banks. Tools considered standard infrastructure for finance professionals in New York or London are suddenly unavailable to their peers in Hong Kong. For investment banks relying on AI to synthesize research, model markets, and automate routine tasks, this disparity directly impacts day-to-day productivity.

The implications for Hong Kong's position as a global financial hub are tangible. The city has historically relied on its seamless integration with Western markets to attract international capital and top-tier talent. If analysts and executives in competing Asian centers like Singapore retain unfettered access to cutting-edge technology, Hong Kong risks becoming structurally disadvantaged. In an industry where marginal efficiencies dictate deal flows, a deficit in foundational tools can influence where banks choose to locate their highest-value work.

This development underscores a fundamental shift in how technology providers and governments view artificial intelligence. AI is no longer operating as a borderless utility. It is increasingly treated as a strategic asset subject to the same geopolitical restrictions that govern advanced semiconductors and sensitive data flows. The "Greater China" carve-out mirrors the compliance frameworks seen in export controls, effectively bringing corporate IT infrastructure into the crosshairs of international tensions.

For investors and corporate executives, the removal of a single AI assistant is less critical than the structural precedent it establishes. It demonstrates that access to mission-critical software can be restricted or withdrawn overnight based purely on geographic location. Financial institutions must now treat technological fragmentation as a core operational risk, requiring them to build resilient, localized tech stacks rather than relying on standardized global deployments.