RPA bots operating without governance create the same risks as ungoverned AI agents: unauthorized actions, no audit trail, no kill switch. Agentomy extends its governance layer to cover bots across any RPA platform or custom automation framework.
Every detection pattern is mapped to a real incident, a real regulatory reference, and a real detection method. The 16-method behavioral monitor runs continuously across the bot lifecycle. No theoretical threats. No generic compliance language. Below are 10 representative patterns from the full 119-pattern library.
Each layer enforces one aspect of RPA governance -- from individual action validation to fleet-wide emergency halt.
Every control mapping references the actual regulatory document. No generic compliance language. All mappings are self-assessed, pending external validation.
| Framework | Controls | Scope |
|---|---|---|
| OCC Bulletin 2023-17 | 4 | Third-party RPA vendor risk assessment, due diligence, ongoing monitoring |
| OCC 2011-12 / SR 11-7 | 4 | Model risk management for AI-enhanced bots. Pure rule-based bots carved out. |
| FFIEC IT Handbook | 3 | Operations monitoring, change management, incident identification |
| DORA Article 6 | 6 | ICT risk management: annual review, board accountability, 4-hour incident reporting |
| EU AI Act Article 6 | 5 | High-risk classification for AI/ML bots in Annex III domains |
| SOX 404 + COSO RPA Framework | 6 | Bot lifecycle ITGCs: development, deployment, monitoring, decommissioning |
| PCI DSS v4.0 | 6 | Bot identity, least privilege, audit logging, tamper detection, incident response |