Hong Kong AI Access Blocked
JPMorgan Restricts Hong Kong Staff from Claude AI Amid US-China Tech Tensions
Key Points
JPMorgan’s move was reported by Financial Times and cited “three people familiar with the matter,” who said the licensing wording triggered the removal of Claude from the bank’s internal “drop-down list” of approved LLMs for Hong Kong staff.
Reuters reported that JPMorgan and Anthropic did not respond to requests for comment, and that Reuters “could not immediately verify” the Financial Times report.
U.S. Commerce Secretary Howard Lutnick’s letter to Anthropic CEO Dario Amodei ordered Anthropic to suspend exports of its Mythos and Fable models “to destinations worldwide and all foreign nationals,” citing concerns they could be used by “military intelligence users in China, Russia and other countries of concern.”
The reports framed Hong Kong as a partial workaround in the AI export landscape: models built by U.S. firms are “not available in mainland China,” but some operate in Hong Kong under limits set by U.S. companies.
JPMorgan Chase has blocked its Hong Kong staff from using Anthropic's Claude AI models, removing them from the bank's internal list of approved tools after a review of the licensing terms, according to Financial Times. The move follows a similar decision by Goldman Sachs in April, when it quietly pulled Claude from its own approved list for Hong Kong bankers.
Reuters said it could not immediately verify the report, and that neither JPMorgan nor Anthropic responded to requests for comment. The back-to-back decisions by two of Wall Street's biggest names signal a growing compliance crackdown on where and how AI tools can be used across Asia.
Licensing Fine Print Triggered the Hong Kong Block
The removal at JPMorgan was not driven by a broad ban on AI tools. Instead, three people familiar with the matter told Financial Times that specific wording in Anthropic's licensing agreement set off alarm bells. New clauses in the terms held the bank liable if Claude was accessed by "foreign nationals of interest." JPMorgan's compliance team decided they could not guarantee that in Hong Kong without legal risk.
Claude was quietly removed from JPMorgan's internal "drop-down list" — the menu staff use to pick approved AI tools. Hong Kong employees can no longer select it. Their counterparts in New York and London still can, according to The Next Web.
Goldman Moved First, JPMorgan Follows a Similar Path
Goldman Sachs took the first step in April, removing Claude from its approved list for Hong Kong bankers in what was described as a "preemptive compliance measure," according to Financial Times. JPMorgan's decision, reported on June 15, puts two of the largest U.S. investment banks on the same side of this issue within a span of two months.
Analysts see a pattern forming. "The cost of compliance is now outweighing the productivity gains of AI for U.S. banks in Asia," Gene Munster of Deepwater Asset Management told Yahoo Finance. An estimated 18,000-plus JPMorgan employees across the Asia-Pacific region are now having their access to AI tools reviewed.
U.S. Government Cracks Down on Anthropic's Frontier Models
The bank-level decisions are playing out against a bigger government push. U.S. Commerce Secretary Howard Lutnick sent a formal letter to Anthropic CEO Dario Amodei ordering the company to suspend exports of its Mythos and Fable models "to destinations worldwide and all foreign nationals." Lutnick cited fears that the models could be used by "military intelligence users in China, Russia, and other countries of concern."
Hong Kong sits in a legal gray zone. U.S. AI models are blocked in mainland China, but some have operated in Hong Kong under limits set by the companies themselves. Since Hong Kong's 2020 National Security Law, U.S. officials have grown less comfortable treating the city as separate from the mainland, according to Guru Focus.
Trump Says Talks With Anthropic Are 'Going Fine'
President Donald Trump addressed the situation at a press briefing on June 17. He said negotiations with Anthropic are "going fine" and added, "we have to be very careful with who gets the best stuff." The comments suggest the White House views the restrictions as leverage in a broader trade and technology negotiation, not a permanent shutdown.
For Anthropic, the pressure is two-sided. The company needs to keep major Wall Street clients happy while also satisfying a U.S. government that views its most powerful models as national security assets. Analysts at Crypto News note that the IPO timeline adds a third dimension — Anthropic must show investors it can manage regulatory risk at scale without losing key commercial relationships.
Sentiment
Mostly NegativePublishers
19
Articles
44
Reach
63
