AI Governance Researcher · Cybersecurity Professional · Founder, NHID-Clinical
I work on operational trust for AI agents — identity, authorization, and auditability for autonomous systems operating in real-world business workflows, with a focus on healthcare administration.
My work grew out of direct exposure to healthcare payer operations, where I identified a practical trust gap: AI voice agents interacting with healthcare organizations without a standardized way to disclose identity or verify authorization.
AI governance increasingly becomes an identity and authorization problem before it becomes a model-performance problem.
I call this gap Impersonation Latency — the time between an AI system initiating an interaction and the receiving party knowing whether that system is authorized to represent the organization it claims to represent. It reframes AI trust as an operational architecture problem rather than a disclosure-only one.
- 🪪 Non-human identity governance for AI agents
- 🛡️ Authorization & deterministic policy engines (identical inputs → identical outputs)
- 📋 Conformance testing — identity disclosure gates, pre-disclosure data prohibition, deceptive behavior checks, escalation controls
- 🔍 Audit-first architecture — structured logs, traceability, replayable evidence
- 🏥 Healthcare AI governance and administrative workflows
- 🗺️ Mapping the AI governance landscape (frameworks, standards, regulation)
| Project | Description |
|---|---|
| NHID-Clinical | Open, voluntary reference implementation for transparency and accountability of AI voice agents in healthcare workflows — specifications, conformance tests, audit-trace schemas, and policy-engine demos. (GitHub org) |
| AI Governance Map | A visualization mapping the fragmented landscape of AI governance frameworks, standards, and regulatory initiatives. |
| Vend-Smart | Applied project exploring smart, agent-driven workflows. |
| Dynamic Media Accessibility & Governance Framework | A governance proposal (v1.3) for safely shipping dynamic/animated profile headers on X — covering WCAG 2.2 photosensitivity compliance, ADA Title II and DSA/EU AI Act alignment, opt-in biometric abuse controls, and a staged rollout plan. |
Overview · Specification · Technical Stack · Evidence Pack · Shadow Evaluation Guide · Interoperability · Regulatory Alignment · For Payers · Roadmap · Developers · Simulator · Community · News
- Are You a Real Person? — Substack
- NIST AI Risk Management comment — public regulatory submission
