Skip to content

Solutions Coverage Gap Analysis

Analysis of FSI-AgentGov-Solutions coverage against the 71-control framework, identifying high-risk gaps and implementation prioritization guidance.


Executive Summary

Metric Value
Total Controls 71
Controls with Deployable Solutions 37
Controls Without Solutions 34
Overall Solution Coverage 52.1%

Important Context

Many "gaps" are addressed by native Microsoft 365 and Power Platform features that require portal configuration, not custom solutions. This analysis focuses specifically on deployable automation from the FSI-AgentGov-Solutions repository. A control without a deployable solution is not necessarily unimplemented—it may be fully addressed through portal configuration following the control's playbooks.


Current Solution Coverage

Covered Controls (37 of 71)

Control ID Control Name Solution
1.1 Restrict Agent Publishing Configuration Hardening Baseline
1.4 Advanced Connector Policies File Upload Security Configurator
1.5 DLP and Sensitivity Labels Deny Event Correlation Report, Scope Drift Monitor
1.7 Comprehensive Audit Logging Deny Event Correlation Report, RAG Source Validator, Configuration Hardening Baseline
1.8 Runtime Protection Deny Event Correlation Report, File Upload Security Configurator, Configuration Hardening Baseline
1.11 Conditional Access and MFA Conditional Access Automation
1.14 Data Minimization Scope Drift Monitor, File Upload Security Configurator
1.18 Application-Level RBAC Configuration Hardening Baseline
2.1 Managed Environments Environment Lifecycle Management, Segregation of Duties Detector, Configuration Hardening Baseline
2.2 Environment Groups Environment Lifecycle Management
2.3 Change Management Message Center Monitor, Pipeline Governance Cleanup, Segregation of Duties Detector
2.4 Business Continuity DR Testing Framework
2.8 Segregation of Duties Segregation of Duties Detector
2.10 Patch Management Message Center Monitor
2.12 Supervision and Oversight FINRA Supervision Workflow, Hallucination Tracker
2.15 Environment Routing Environment Lifecycle Management
2.16 RAG Source Integrity RAG Source Validator
2.18 Conflict of Interest Testing COI Testing Framework
3.3 Compliance Reporting Compliance Dashboard
3.4 Incident Reporting Deny Event Correlation Report
3.7 PPAC Security Posture Configuration Hardening Baseline
3.8 Copilot Hub Agent Access Governance Monitor, Configuration Hardening Baseline
3.10 Hallucination Feedback Hallucination Tracker
1.9 Data Retention and Deletion Policies DR Testing Framework
1.10 Communication Compliance Monitoring FINRA Supervision Workflow, MIME Type Restrictions
1.13 Sensitive Information Types (SITs) MIME Type Restrictions
1.23 Step-Up Authentication Session Security Configurator, Conditional Access Automation
1.25 MIME Type Restrictions for File Uploads MIME Type Restrictions
2.5 Testing, Validation, and Quality Assurance COI Testing Framework
2.9 Agent Performance Monitoring Hallucination Tracker, Agent Usage & Performance Workbook
2.11 Bias Testing and Fairness Assessment COI Testing Framework
2.13 Documentation and Record Keeping RAG Source Validator
2.22 Inactivity Timeout Enforcement Configuration Hardening Baseline, Inactivity Timeout Enforcement
3.1 Agent Inventory and Metadata Management Compliance Dashboard
3.2 Usage Analytics and Activity Monitoring Compliance Dashboard, Agent Usage & Performance Workbook
3.9 Microsoft Sentinel Integration Agent Usage & Performance Workbook
4.3 Site and Document Retention Management MIME Type Restrictions

Coverage by Pillar

Pillar Total Controls Covered Gaps Coverage
Pillar 1 - Security 28 13 15 46.4%
Pillar 2 - Management 24 15 9 62.5%
Pillar 3 - Reporting 12 8 4 66.7%
Pillar 4 - SharePoint 7 1 6 14.3%
Total 71 37 34 52.1%

Gap Classification

Not all gaps require custom solutions. Understanding the gap type helps prioritize development efforts.

Category 1: Native Microsoft Features (Portal Configuration)

These controls are fully addressed through Microsoft admin portal configuration. No custom solution is needed—follow the control's playbooks for implementation.

Control Control Name Native Feature
~~1.1~~ ~~Restrict Agent Publishing~~ ~~Power Platform Admin Center settings~~ Now also has Configuration Hardening Baseline
1.2 Agent Registry Integrated Apps management in M365 Admin
~~1.4~~ ~~Advanced Connector Policies~~ ~~PPAC connector classification~~ Now has File Upload Security Configurator
1.6 DSPM for AI Purview compliance portal
~~1.8~~ ~~Runtime Protection~~ ~~Defender for Cloud Apps~~ Now has Deny Event Correlation Report and File Upload Security Configurator
1.9 Data Retention Purview retention policies
1.10 Communication Compliance Purview Communication Compliance
~~1.11~~ ~~Conditional Access~~ ~~Entra ID Conditional Access policies~~ Now has Conditional Access Automation
1.12 Insider Risk Detection Purview Insider Risk Management
1.13 Sensitive Information Types Purview SIT configuration
1.15 Encryption Microsoft-managed encryption (default)
1.16 Information Rights Management Purview sensitivity labels with IRM
1.17 Endpoint DLP Purview Endpoint DLP
~~1.18~~ ~~Application-Level RBAC~~ ~~Entra ID app roles~~ Now also has Configuration Hardening Baseline
1.19 eDiscovery Purview eDiscovery
1.22 Information Barriers Purview Information Barriers
1.24 Defender AI-SPM Defender for Cloud
3.1 Agent Inventory PPAC + CoE Starter Kit
3.2 Usage Analytics PPAC Analytics
~~3.7~~ ~~PPAC Security Posture~~ ~~PPAC native dashboard~~ Now also has Configuration Hardening Baseline
~~3.8~~ ~~Copilot Hub~~ ~~M365 Admin Center Copilot Hub~~ Now has Agent Access Governance Monitor
4.1-4.7 SharePoint Controls SharePoint Advanced Management

These controls would benefit from custom automation beyond native features. Prioritize based on Zone 3 / Priority Tier 1 risk.

Control Control Name Gap Type Priority
1.3 SharePoint Content Governance Automated permission scanning Medium
~~1.14~~ ~~Data Minimization~~ ~~Scope drift detection automation~~ ADDRESSED
1.20 Network Isolation Connectivity validation automation Medium
1.21 Adversarial Input Logging Centralized attack pattern analysis High
1.23 Step-Up Authentication Auth challenge orchestration Medium
~~2.4~~ ~~Business Continuity~~ ~~Automated DR testing~~ ADDRESSED
2.5 Testing and Validation Test orchestration framework High
~~2.8~~ ~~Segregation of Duties~~ ~~Role conflict detection~~ ADDRESSED
2.9 Performance Monitoring Custom KPI dashboards Medium
~~2.16~~ ~~RAG Source Integrity~~ ~~Source validation automation~~ ADDRESSED
2.17 Multi-Agent Orchestration Orchestration limit enforcement Medium
~~2.18~~ ~~Conflict of Interest Testing~~ ~~Automated COI detection~~ ADDRESSED
~~3.3~~ ~~Compliance Reporting~~ ~~Aggregated compliance dashboard~~ ADDRESSED
3.5 Cost Allocation Chargeback automation Medium
3.6 Orphaned Agent Detection Automated remediation workflows Medium
3.9 Sentinel Integration Custom data connectors High
~~3.10~~ ~~Hallucination Feedback~~ ~~Feedback aggregation pipeline~~ ADDRESSED

Category 3: Process/Documentation Controls

These controls are inherently process-based and cannot be fully automated. They require documented procedures, training, and human judgment.

Control Control Name Why Not Automatable
2.6 Model Risk Management OCC 2011-12 requires human model validation and governance committee oversight
2.7 Vendor Risk Management Third-party assessments require human judgment and negotiation
2.11 Bias Testing Fairness assessment requires domain expertise and ethical judgment
2.12 Supervision and Oversight ~~FINRA 3110 requires designated supervisory principals~~ Now has FINRA Supervision Workflow - Automates routing and tracking while human review remains required
2.13 Documentation Record-keeping is a process discipline
2.14 Training Program Human learning and awareness
2.19 AI Disclosure Customer communication decisions
2.20 Adversarial Testing Red team exercises require human creativity
2.21 AI Marketing Claims Legal/compliance review of marketing materials

High-Risk Gaps (Zone 3 / Priority Tier 1)

These 32 controls apply to Zone 3 (Enterprise Managed) agents and Priority Tier 1 (Critical) materiality classifications. Gaps in these controls present elevated regulatory and operational risk.

Zone 3 Controls Without Solutions

Control Control Name Gap Category Regulatory Impact
1.3 SharePoint Content Governance Custom Recommended Data exposure risk
~~1.4~~ ~~Advanced Connector Policies~~ ~~Native Feature~~ ADDRESSED
1.6 DSPM for AI Native Feature -
~~1.8~~ ~~Runtime Protection~~ ~~Native Feature~~ ADDRESSED
1.9 Data Retention Native Feature SEC 17a-4, FINRA 4511
1.10 Communication Compliance Native Feature FINRA 3110
~~1.11~~ ~~Conditional Access~~ ~~Native Feature~~ ADDRESSED
1.12 Insider Risk Detection Native Feature -
~~1.14~~ ~~Data Minimization~~ ~~Custom Recommended~~ ADDRESSED
1.19 eDiscovery Native Feature SEC 17a-4, FINRA 4511
1.20 Network Isolation Custom Recommended Security posture
1.21 Adversarial Input Logging Custom Recommended Threat detection
1.22 Information Barriers Native Feature FINRA Rule 5110
1.23 Step-Up Authentication Custom Recommended Access control
1.24 Defender AI-SPM Native Feature -
~~2.4~~ ~~Business Continuity~~ ~~Custom Recommended~~ ADDRESSED
2.5 Testing and Validation Custom Recommended FINRA Notice 15-09
2.6 Model Risk Management Process Control OCC 2011-12, SR 11-7
2.7 Vendor Risk Management Process Control OCC 2013-29
~~2.8~~ ~~Segregation of Duties~~ ~~Custom Recommended~~ ADDRESSED
2.9 Performance Monitoring Custom Recommended SLA compliance
2.11 Bias Testing Process Control ECOA, CFPB
2.13 Documentation Process Control SEC 17a-3
~~2.16~~ ~~RAG Source Integrity~~ ~~Custom Recommended~~ ADDRESSED
2.17 Multi-Agent Orchestration Custom Recommended Complexity risk
~~2.18~~ ~~Conflict of Interest Testing~~ ~~Custom Recommended~~ ADDRESSED
2.19 AI Disclosure Process Control CFPB guidance
2.20 Adversarial Testing Process Control Security posture
~~3.3~~ ~~Compliance Reporting~~ ~~Custom Recommended~~ ADDRESSED
3.9 Sentinel Integration Custom Recommended SIEM coverage
~~3.10~~ ~~Hallucination Feedback~~ ~~Custom Recommended~~ ADDRESSED

Critical Regulatory Gaps

One control has significant regulatory implications with no native Microsoft feature fully addressing it.

FINRA Rule 3110 - Supervision (Control 2.12) - ADDRESSED

Solution Available

The FINRA Supervision Workflow solution provides automated routing and tracking for supervisory review while maintaining human oversight as required by regulation.

Regulation: FINRA Rule 3110 requires member firms to establish and maintain a system to supervise the activities of each associated person that is reasonably designed to achieve compliance.

Solution: The FINRA Supervision Workflow solution automates:

  • Queue management for flagged AI agent outputs
  • Routing to designated supervisory principals based on zone/tier
  • SLA monitoring with automatic escalation
  • Evidence collection with SHA-256 integrity hashing
  • Integration with Communication Compliance for content detection

Remaining Human Requirements:

  1. Designate supervisory principals with AI agent oversight responsibilities
  2. Document supervision procedures in Written Supervisory Procedures (WSP)
  3. Perform actual content review and approval decisions
  4. Configure zone/tier-specific review percentages and SLAs

See: Control 2.12 Playbooks

OCC 2011-12 / SR 11-7 - Model Risk Management (Control 2.6)

Regulation: OCC Bulletin 2011-12 and Fed SR 11-7 require banks to establish a model risk management framework including model validation, ongoing monitoring, and governance.

Gap: Microsoft provides infrastructure for deploying AI agents, not a pre-built model risk management solution. MRM requires human judgment and governance structures.

Mitigation:

  1. Document AI agents in your model inventory
  2. Classify agents by materiality tier (Tier 1/2/3)
  3. Establish model validation procedures appropriate to agent complexity
  4. Define model owner and validator roles with appropriate independence
  5. Implement ongoing performance monitoring (Control 2.9)
  6. Maintain model documentation per SR 11-7 requirements

See: Control 2.6 Playbooks


Implementation Roadmap Priority

Recommended prioritization for addressing solution gaps, organized by quarter. Effort estimates assume a dedicated implementation team with Power Platform and Azure experience.

Phase 1: Q1 2026 (Foundation)

Focus on controls with highest regulatory impact and broadest applicability.

Priority Control Action Approach Effort
1 2.12 ~~Configure supervision workflow~~ DONE - Deploy FINRA Supervision Workflow Solution Available -
2 2.6 Establish MRM governance for AI agents; create model inventory Process + Documentation 3 weeks
3 1.22 Configure Information Barriers for research/trading separation Portal Configuration 1 week
4 1.11 ~~Deploy Conditional Access policies for AI app access~~ DONE - Deploy Conditional Access Automation Solution Available -
5 3.3 ~~Build compliance reporting dashboard using existing audit data~~ DONE - Deploy Compliance Dashboard Solution Available -

Phase 2: Q2 2026 (Hardening)

Extend coverage to operational excellence controls.

Priority Control Action Approach Effort
1 2.5 Implement automated testing framework for Zone 3 agents Custom Development 6 weeks
2 2.8 ~~Build role conflict detection for Maker/Checker enforcement~~ DONE - Deploy Segregation of Duties Detector Solution Available -
3 1.14 ~~Deploy scope drift detection automation~~ DONE - Deploy Scope Drift Monitor Solution Available -
4 3.9 Configure Sentinel MCP Server for agent telemetry Portal + Integration 2 weeks
5 2.16 ~~Implement RAG source validation checks~~ DONE - Deploy RAG Source Validator Solution Available -

Phase 3: Q3 2026 (Optimization)

Address remaining gaps and enhance automation maturity.

Priority Control Action Approach Effort
1 2.18 ~~Deploy automated conflict of interest testing~~ DONE - Deploy COI Testing Framework Solution Available -
2 1.21 Centralize adversarial input analysis Custom Development 3 weeks
3 3.10 ~~Build hallucination feedback aggregation pipeline~~ DONE - Deploy Hallucination Tracker Solution Available -
4 2.4 ~~Implement automated DR testing workflows~~ DONE - Deploy DR Testing Framework Solution Available -
5 3.5 Deploy cost allocation and chargeback automation Custom Development 3 weeks

Phase 4: Q4 2026 (Maturity)

Achieve comprehensive automation coverage and operational excellence.

Priority Control Action Approach Effort
1 1.3 Automated SharePoint permission scanning Custom Development 4 weeks
2 1.20 Network isolation connectivity validation Custom Development 3 weeks
3 2.17 Multi-agent orchestration limit enforcement Custom Development 4 weeks
4 3.6 Orphaned agent automated remediation Custom Development 3 weeks
5 1.23 Step-up authentication challenge orchestration Custom Development 4 weeks

Solution Development Backlog

Priority solutions for FSI-AgentGov-Solutions repository development, addressing critical regulatory and operational gaps.

P0 - Critical (Q1-Q2 2026) - ALL RELEASED ✓

Solution Target Control Description Status
~~finra-supervision-workflow~~ 2.12 ~~Automated supervision queue for AI agent outputs~~ RELEASED v1.0.0
~~conditional-access-automation~~ 1.11 ~~Entra ID Conditional Access policy templates and deployment automation for AI workloads~~ RELEASED v1.0.0
~~compliance-dashboard~~ 3.3 ~~Aggregated compliance reporting across all 71 controls with zone-based filtering~~ RELEASED v1.0.0

P1 - High (Q2-Q3 2026) - ALL RELEASED ✓

Solution Target Control Description Status
~~segregation-detector~~ 2.8 ~~Role conflict detection for Maker/Checker enforcement in agent pipelines~~ RELEASED v1.0.0
~~scope-drift-monitor~~ 1.14 ~~Automated detection of agent data access beyond declared scope~~ RELEASED v1.0.0
~~rag-source-validator~~ 2.16 ~~Integrity validation for RAG knowledge sources with change detection~~ RELEASED v1.0.0

P2 - Medium (Q3-Q4 2026) - ALL RELEASED ✓

Solution Target Control Description Status
~~coi-testing~~ 2.18 ~~Automated conflict of interest testing for agent recommendations~~ RELEASED v1.0.0
~~hallucination-tracker~~ 3.10 ~~Feedback aggregation pipeline for hallucination detection patterns~~ RELEASED v1.0.0
~~dr-testing-framework~~ 2.4 ~~Automated disaster recovery testing for agent infrastructure~~ RELEASED v1.0.0

Duplicate Coverage Analysis

Some controls are covered by multiple solutions, providing redundancy.

Control 2.3 - Change Management

Solution Coverage Aspect
Message Center Monitor Platform change notifications from Microsoft
Pipeline Governance Cleanup Internal deployment pipeline governance

Recommendation: Deploy both solutions. They address different aspects of change management—external platform changes vs. internal deployment governance.

Controls 1.5, 1.7, and 1.8

Solution Controls
Deny Event Correlation Report 1.5 (DLP), 1.7 (Audit), 1.8 (Runtime Protection), 3.4 (Incident)

Note: The Deny Event Correlation Report provides unified visibility across multiple controls by correlating events from Purview Audit, DLP, Defender for Cloud Apps, and Application Insights.

Controls 1.4, 1.8, and 1.14

Solution Controls
File Upload Security Configurator 1.4 (Connector Policies), 1.8 (Runtime Protection), 1.14 (Data Minimization)

Note: The File Upload Security Configurator validates per-agent file upload settings against zone governance policies, covering connector-level restrictions, runtime threat detection, and data minimization controls.


Next Steps

  1. Assess current state - For each gap, determine if native features are already configured
  2. Prioritize by zone - Focus on Zone 3 gaps first for enterprise agents
  3. Leverage existing solutions - Deploy available FSI-AgentGov-Solutions before building custom
  4. Document process controls - For Category 3 controls, focus on procedures and training
  5. Track progress - Use the implementation roadmap to measure gap closure


FSI Agent Governance Framework v1.2.51 - February 2026