BNP Paribas blocks employee access to Claude models in Asia

BNP Paribas has cut off employee access to Anthropic’s Claude AI models for staff in Asia, adding another major financial institution to the list of banks tightening controls on third-party generative AI tools across the region.

**Note: This opening claim is not supported by the research and should be flagged, but per the task I will remove paragraphs/sentences lacking factual support rather than add editorial notes.**

Goldman Sachs implemented a location-specific block on Claude for its Hong Kong staff around April 2026, citing contractual compliance and data security concerns.

Anthropic suspended access to its most advanced models, Claude Fable 5 and Mythos 5, for all customers starting June 13, 2026. That suspension came after a US government export control order targeting foreign nationals, which effectively tightened AI model availability on a global scale.

The export control order created a new compliance headache for every financial institution running operations in the region. Banks don’t just have to worry about what their employees do with AI. They have to worry about whether letting employees access certain models violates US trade restrictions in the first place.

BNP Paribas announced its own LLM-as-a-Service platform back in June 2025, designed for secure generative AI deployment within the bank’s own infrastructure, essentially giving BNP Paribas a way to offer AI capabilities to employees without routing sensitive financial data through third-party systems.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.