JPMorgan cuts off Anthropic AI access for Hong Kong staff amid tightening US export controls

JPMorgan Cuts Off Anthropic Access Over Licensing Concerns

JPMorgan has removed Anthropic's Claude models from the internal list of approved large language models available to employees in Hong Kong, according to reports from the Financial Times 1

1

. The decision centers on the wording of Anthropic's usage terms in its licensing agreement with JPMorgan, creating compliance concerns in one of the world's most geopolitically sensitive financial hubs 4

4

. Staff in Hong Kong can no longer select Claude from the drop-down menu of AI tools their colleagues elsewhere continue to use, marking a jurisdictional restriction rather than a verdict on the model's performance 1

1

.

US Banking Giants Restrict AI Following Similar Pattern

This marks the second time a major Wall Street institution has pulled Claude from Hong Kong operations. Goldman Sachs removed Claude from its list of approved tools available to Hong Kong-based bankers in April, adopting a strict interpretation of usage rules related to Greater China 2

2

. Two of the largest US banks reaching the same conclusion about the same vendor in the same city suggests an emerging industry norm driven by AI governance considerations 1

1

. Both institutions sent 'Do not use' notes on Anthropic to employees outside the US, highlighting how AI model access restrictions are reshaping corporate technology policies 2

2

.

US National Security Crackdown Drives Export Controls

The restrictions arrive amid intensifying political pressure from Washington over advanced AI models. US Commerce Secretary Howard Lutnick ordered Anthropic to suspend exports of its most advanced models, including Mythos and Fable, to foreign nationals, citing risks that they could be diverted to military or intelligence users in China, Russia, and other countries of concern 1

1

4

4

. The Commerce Department directive reflects escalating U.S.-China tensions over AI technology, data security, and access to advanced computing tools 4

4

. For banks, Hong Kong sits awkwardly in this landscape as both a Chinese territory and global financial center, making any tool whose licensing terms create ambiguity about cross-border data a compliance question rather than a productivity one 1

1

.

Geopolitical and Corporate Implications of AI Access

Anthropic has separately barred its products from use in mainland China, concerned that Chinese AI firms could train their own models through distillation attacks 3

3

. While mainland China blocks Western AI tools like ChatGPT and Claude under the country's Great Firewall, Hong Kong has traditionally maintained broader access to global digital platforms 2

2

. The timing proves awkward for Anthropic, which has lined up Morgan Stanley and Goldman Sachs to lead its IPO, with JPMorgan also reported to be working on the deal 1

1

. The same banks selling Anthropic to public markets are restricting its use in certain offices, creating a visible contradiction as the company courts investors at a reported valuation in the hundreds of billions 1

1

.

Europe Pushes for Sovereign AI Capabilities

The US move has sparked concern in Europe, where policymakers fear over-reliance on American AI infrastructure. French President Emmanuel Macron has led discussions with G7 leaders and tech executives on ensuring access to advanced AI models through trusted partnerships, while German Chancellor Friedrich Merz emphasized the need for coordination with Washington on making new technologies available for all countries 2

2

. European leaders are increasingly pushing for sovereign AI to reduce dependency on US technology firms 2

2

. Access to frontier models is becoming increasingly dependent on geography, regulation, and security assessments, reshaping how global AI adoption unfolds and creating a widening fault line in the technology sector 1

1

.