ICT risk management, third-party oversight — AI agent registry and vendor access controls directly apply.
This page is a working reference and not a substitute for qualified legal review. Verify against official sources before use in compliance artifacts.
The policy below implements DORA's ICT risk management requirements for AI agents accessing financial data. Agent registry + source registry together satisfy the ICT asset inventory DORA Articles 5-16 require.
# FS-DORA-ART5-001 — ICT Risk Management via Agent Registry
# DORA Articles 5-16: ICT risk framework + third-party oversight
# Applies from 17 January 2025 — no phase-in period
- name: dora_ict_governance_agent
policy_id: FS-DORA-ART5-001
agent_role: dora_ict_agent
allowed_sources:
- ict_asset_registry
- third_party_contracts
- incident_reports
- resilience_test_results
- vendor_assessments
denied_sources:
- customer_financial_records
- payment_transaction_data
- trading_positions
- client_portfolios
allowed_tasks:
- ict_inventory_review
- third_party_risk_assessment
- incident_classification
- resilience_testing_review
max_sensitivity: high
require_task_for_sensitivity: high
# DORA Art. 28: third-party risk — each task bound to minimum required sources
task_bindings:
- task: ict_inventory_review
permitted_sources: [ict_asset_registry, third_party_contracts]
- task: third_party_risk_assessment
permitted_sources: [vendor_assessments, third_party_contracts, ict_asset_registry]
- task: incident_classification
permitted_sources: [incident_reports, ict_asset_registry]
- task: resilience_testing_review
permitted_sources: [resilience_test_results, ict_asset_registry]
session_ttl_minutes: 480
require_principal_entitlements:
- tasks: [third_party_risk_assessment]
require_any: [group:ict-risk-officers, group:ciso-team]
DORA applies to financial entities regulated in the EU and their ICT third-party providers. AI systems that access, process, or generate outputs from financial data are ICT assets under DORA's scope.
AutoPIL intercepts every AI agent data access call, enforces your policy, and writes a tamper-evident audit record — before sensitive data enters the agent context window.