JPMorgan, Goldman Sachs Ban Hong Kong Staff from Claude AI Amidst Global Regulatory Pressures
June 18, 2026
JPMorgan Chase and Goldman Sachs have barred Hong Kong employees from accessing Anthropic’s Claude, reflecting heightened scrutiny of AI tools outside the US and a shift toward more restricted, enterprise-grade solutions.
The restriction stems from Anthropic’s licensing terms and removes Claude from JPMorgan’s list of approved large language models, signaling a contractual/permissions-based containment rather than performance concerns.
The move sits within wider regulatory pressure, including a U.S. Commerce Department request to suspend Mythos and Fable exports to certain destinations and nationals over potential military-intelligence risks.
Industry observers say the development may spur providers to offer more secure, governable enterprise AI options and could reshape how finance teams adopt external AI tools.
Analysts expect AI deployment in banking to proceed unevenly by region, guided by licensing terms, governance standards, and geopolitical factors.
Regulators globally are tightening expectations for AI in financial services, pressing banks to strengthen governance, data usage controls, model audits, and transparency.
Despite these restrictions, AI remains a competitive factor in trading, risk management, and operations, pushing banks to balance innovation with compliance and security.
The trend toward private, on-premise AI deployments may accelerate in regulated sectors as firms seek greater data control and security.
This shift aligns with a broader push for robust AI governance, with stricter oversight and a move away from external cloud-based models in regulated industries.
International discussions, including at the G7, emphasize balancing AI safety with broad accessibility and cross-border collaboration on trusted AI partnerships.
Industry leaders and policymakers stress the need to maintain safe, accessible AI while managing safety concerns in a global context.
Analysts view the Hong Kong move as a board-level signal that banks may regionalize AI tooling, potentially prompting other US and European banks to follow JPMorgan’s lead.
Summary based on 20 sources
Get a daily email with more Tech stories
Sources

Economic Times • Jun 18, 2026
'Do not use': US banking giants JPMorgan and Goldman Sachs restrict Anthropic AI access abroad. Here's why
The Times Of India • Jun 18, 2026
JPMorgan Chase and Goldman Sachs send 'Anthropic ban' message to employees outside the US; say: Stop ...
The Next Web • Jun 18, 2026
JPMorgan cuts off Anthropic access for Hong Kong staff
Blockonomi • Jun 18, 2026
JPMorgan Cuts Claude AI Access in Hong Kong Amid Rising Security Concerns