Trackers · Living reference

AI Enterprise-Ban & Export-Control Tracker

Who has restricted, banned, or export-controlled which AI vendor — and where. A running record of the moment frontier-model access stopped being a vendor question and became a compliance and policy one. Updated as events occur.

Last updated 2026-06-20 · 4 events

DateOrganizationActionVendor / modelJurisdictionSource
Jun 18, 2026JPMorgan ChaseEnterprise restrictionRemoved Claude from the internal approved-LLM drop-down for Hong Kong employees, citing the wording of Anthropic’s licensing terms.Anthropic (Claude)Hong KongReuters / FTca.finance.yahoo.com
Jun 14, 2026U.S. Commerce DepartmentGovernment export controlExport-control directive banned Fable 5 and Mythos 5 for use by any foreign national, inside or outside the U.S.; Anthropic disabled both models worldwide to comply.Anthropic (Fable 5, Mythos 5)U.S. export control (global effect)Fox Business / Anthropicfoxbusiness.com
Jun 12, 2026U.S. bank regulators (OCC + Federal Reserve)Regulatory scrutinyBegan probing how banks deploy AI in routine exams — pressing on human "kill switches", data access, and third-party vendor risk.All third-party AI vendorsUnited States (banking)Reutersaol.com
Apr 2026Goldman SachsEnterprise restrictionRemoved Claude from the list of approved tools available to its Hong Kong–based bankers — weeks ahead of JPMorgan’s similar move.Anthropic (Claude)Hong KongReuters / FTca.finance.yahoo.com

How this is compiled

  • Every row cites a named primary source. Events asserted only in commentary, without a primary link, are left out until they’re sourced.
  • Three categories: enterprise restriction (a company limits a vendor internally), government export control (a state restricts a model’s availability), and regulatory scrutiny (supervisors pressure-test AI deployments).
  • Drawn from Agentic Daily’s reporting. Spot something missing or out of date? Tell us.

Informational only · not legal or compliance advice