io.github.uchit/mcp-regulated-ai-compliance

Assess the compliance risks of this AI customer service system for the financial sector under the EU AI Act, and list remediation recommendations by high, medium, and low priority:
- Functions: answer customer questions automatically, generate investment product explanations
- Data: customer chat history, account information
- Region: EU market
- Current controls: manual spot checks, basic logging

A structured risk assessment with identified issues, applicable obligations, and prioritized remediation actions.

Map our existing generative AI governance measures to the NIST AI RMF and ISO 42001. Identify covered, partially covered, and missing controls, and recommend how to close the gaps. Existing measures include: model approval workflow, prompt filtering, human review of outputs, access controls, and incident reporting.

A control mapping and gap analysis showing which framework requirements are covered and what still needs improvement.

We are preparing to launch an AI system for insurance claims triage in Australia. Using APRA and AU AI Safety requirements, generate a review-readiness checklist covering governance, risk management, documentation retention, monitoring, and accountability.

An actionable compliance checklist the team can use to verify launch readiness item by item.