JPMorgan blocks Anthropic Claude for staff in Hong Kong

In short: JPMorgan Chase has stopped employees in Hong Kong from accessing Anthropic’s Claude AI models through the bank’s internal approved tools.

What happened

JPMorgan Chase has cut off access to Anthropic’s Claude for its Hong Kong staff, according to the Financial Times, which cited three people familiar with the situation.

The change affects an internal list of approved AI tools that employees can pick from. Staff in Hong Kong can no longer select Claude from that list.

One person told the FT the decision was linked to the wording in Anthropic’s licensing terms with JPMorgan. The report says Goldman Sachs made a similar move earlier this year, after reading Anthropic’s terms as excluding use in “Greater China,” which includes Hong Kong.

Claude and ChatGPT are not available in mainland China, due to China’s internet controls. Hong Kong has typically had fewer restrictions, but in this case the limits are being set by US AI companies and contract terms, not by Hong Kong authorities.

US AI companies have also raised concerns about their models being used in China, including the risk of “distillation,” which is when someone tries to build a new AI system by studying the outputs of an existing one (like learning a recipe by repeatedly tasting a dish).

Why it matters

Big banks and other employers are still deciding when and where staff can use popular AI tools, especially across different countries. For workers, this can change what tools they can use for everyday tasks like writing, research, and coding, and it can create uneven access between offices in different regions.

Source: Financial Times