Inventor — Pharns Genece — Dual-Domain Governance Architecture¶
Pharns Genece is the inventor of the governed execution framework — a runtime AI agent governance system that enforces policy at dispatch, before AI agents act, not after. Five USPTO provisional patent applications filed April–May 2026. NIST OLIR listed. Operational today, governing its inventor's own multi-agent system.
My perspective comes from working across the autonomy continuum: automation → autonomy → AI. It started in the initial cohort of MiraCosta College's ROV/Drone Operator Technician Program — remotely operated underwater vehicles and unmanned aircraft, machine design and FAA regulation — which I returned to teach in 2019. From there: physical autonomous systems, RF and spectrum sensing, then cybersecurity and digital AI agents. I governed machines that could cause great harm before I governed software, and the control problem turned out to be the same one.
The Patent Record¶
First Provisional Application Filed: April 4, 2026 Scope: AI agent governance foundation — dispatch-time enforcement, risk-tiered classification (R1/R2/R3), trust-weighted governed memory, multi-domain RF sensing (TraceLock embodiment).
Second Provisional Application Filed: April 25, 2026 Scope: Modular governance architecture, edge subordinate enforcement, governed model selection (C1/C2/C3 complexity tiers), sensor-actuated AI dispatch, AAM governance, AQ-based epistemic independence measurement.
Third Provisional Application Filed: May 16, 2026 Title: Governed Autonomous Execution System with Structural Isolation, Formal Enforcement Invariance, Governance Lifecycle Properties, and Demonstrated Resilience. Scope: structural isolation and formal guarantees, governance lifecycle, adversarial hardening, principal hierarchy (commercial / AQ Scale measurement standard), and OLIR gap closure.
Fourth Provisional Application Filed: May 19, 2026 Title: Governed Persistent Orchestration Identity for AI Agent Runtimes Across Sessions and Modalities. Scope: Governed Persistent Orchestration Identity (GPOI) — identity continuity for AI agent runtimes across sessions and modalities. Governance gateway holds, retrieves, and verifies Identity Continuity Token; agent presents identifier only. Fail-closed enforcement when verification infrastructure is unavailable.
Fifth Provisional Application Filed: May 28, 2026 Title: Topology-Adaptive Governed Subordinate Runtime with Process-Boundary Enforcement, Combined-Signal Detection, Pre-Action Memory, and Always-On Governance. Scope: Topology-adaptive governed subordinate runtime — runtime selection of governance topology with per-topology integrity attestation, process-boundary enforcement, combined-signal covert-channel detection, pre-action memory verification, and a fail-closed always-on governance runtime.
Combined scope: 45 patent families | 334 total claims across five provisionals.
Filing receipts are available to editors, counsel, and diligence reviewers on request.
Non-provisional filing target: April 4, 2027 (Prov #1 expiry — strict-conservative posture preserves all five priority dates).
What Was Invented¶
Most AI governance frameworks operate at the policy layer — they produce documents, audit logs, and compliance reports. The gap they leave is the moment that matters: the instant an AI agent acts.
The governed execution framework addresses that gap directly.
The governed execution framework is a runtime AI agent governance system that classifies risk, enforces policy, and bounds authority at dispatch — before the agent acts, not after. Three NIST OLIR Concept Crosswalks cataloged: AI RMF 1.0 (Ref ID 220), CSF 2.0 (Ref ID 215), and SP 800-53 Rev 5.2.0 (Ref ID 217) — all Final Informative References.
What dispatch-time enforcement means: Every agent action is intercepted before execution, evaluated against a deterministic rule set, and issued a verdict — PERMIT, DENY, or CONDITIONAL. No verdict, no action. The agent cannot override it.
Core inventions:
- Dispatch-time enforcement — policy applied at the moment of agent decision, not post-hoc
- Risk-tiered classification — R1/R2/R3 risk tiers govern which agents can act and under what conditions
- Governed model selection — complexity tiers (C1/C2/C3) route decisions to appropriately capable models; agents cannot self-promote
- Trust-weighted governed memory — retrieval is a governed operation with audit trails
- Pre-egress policy enforcement — covert-channel detection before any agent output leaves the system
- Multi-domain RF sensing — RF + behavioral + network fusion as a governance input (TraceLock embodiment)
- Modular encrypted governance — signed, hot-swappable governance modules with cryptographic integrity verification
- Structural execution-governance separation — governance state architecturally isolated from any execution path; no agent can write to or modify the policy under which it operates
- Formal enforcement invariance — governance properties verified to hold across all reachable system states, including adversarial conditions
- Behavioral envelope enforcement — each agent declares an authorized action envelope at admission; runtime divergence triggers escalation before the action completes
Public Systems¶
The following implementations are cleared for public reference:
| System | Repository | Disclosure Level |
|---|---|---|
| Governed execution framework — core governance | sdos-public | Tier 2 — what it does |
| TraceLock — RF detection | tracelock-public | Tier 2 — what it does |
| GIAP — governance intake | giap-public | Tier 2 — what it does |
| Privacy-first email | privacy-first-email-public | Tier 2 — what it does |
All systems are covered by filed U.S. provisional patent applications. Implementation details beyond "what it does" are not disclosed pending non-provisional filing.
External Validation¶
NIST OLIR Listing Three NIST OLIR Concept Crosswalks cataloged (OLIR Trifecta): - Reference ID 220 — SDOS-RuntimeGov-to-AI-RMF-v1.0 (AI RMF 1.0, Final — public review concluded 2026-06-15) - Reference ID 215 — SDOS-RuntimeGov-to-CSF-2.0-v1.0 (CSF 2.0, Final — public review concluded 2026-06-22) - Reference ID 217 — SDOS-RuntimeGov-to-SP-800-53-Rev-5.2.0-v1.0 (SP 800-53 Rev 5.2.0, Final — public review concluded 2026-06-22)
Three Five Laws of AI Governance crosswalks catalogued as Draft Informative References (2026-10-06, in 30-day public review): - FiveLaws-AIGov-to-AI-RMF-v1.0 (AI RMF 1.0, Draft) - FiveLaws-AIGov-to-CSF-v2.0 (CSF 2.0, Draft) - FiveLaws-AIGov-to-SP800-53r5 (SP 800-53 Rev 5.2.0, Draft)
NIST OLIR is a federal catalog of cybersecurity and AI governance reference frameworks. An OLIR entry is a mapping authored by the submitting organization, reviewed for completeness against the OLIR program's own criteria, published on a federal site, and opened for public comment. All three of these reached Final status with zero comments received. Listing is an informative reference, not a NIST endorsement, certification, or assessment of the framework itself.
Published Reference Pages The governed execution framework is publicly documented with framework alignment mappings for 17 regulatory standards including NIST AI RMF, NIST AI 600-1, ISO 42001, EU AI Act, FedRAMP, CMMC, HIPAA, and NERC CIP. The reference documentation covers the full governance architecture — dispatch-time enforcement model, risk-tier classification, and control mappings.
Reference Documentation — the technical architecture behind the patent claims.
Why This Matters¶
The AI governance market is producing frameworks. What it is not producing is enforcement.
The distinction is architectural. A framework tells you what to do. An enforcement system does it — at runtime, under policy, with an audit trail.
That is the gap the patents address. That is the problem these systems were built to solve.
Frequently Asked Questions¶
Who invented the governed execution framework? Pharns Genece, founder of AAM Cyber, invented the governed execution framework. The system is covered by five U.S. provisional patent applications filed in April and May 2026.
What does dispatch-time enforcement mean? It means policy is enforced at the moment an AI agent decides to act — before the action executes — not audited after the fact. Every action is intercepted, evaluated against a deterministic rule set, and issued a verdict: PERMIT, DENY, or CONDITIONAL.
What is the patent status of the governed execution framework? Five USPTO provisional patent applications are filed, dated April 4, April 25, May 16, May 19, and May 28, 2026, covering 45 patent families and 334 total claims. Filing receipts are available to editors, counsel, and diligence reviewers on request. Non-provisional consolidating all five is in preparation with a target filing of April 4, 2027 (Prov #1 expiry — strict-conservative posture preserves all five priority dates).
Is the governed execution framework listed in any federal catalog? Three NIST OLIR Concept Crosswalks are cataloged — the OLIR Trifecta: Reference ID 220 (AI RMF 1.0, Final), Reference ID 215 (CSF 2.0, Final), and Reference ID 217 (SP 800-53 Rev 5.2.0, Final). All three are Final Informative References (public review concluded with zero comments: AI RMF 1.0 on 2026-06-15, CSF 2.0 and SP 800-53 Rev 5.2.0 on 2026-06-22). Three further crosswalks, mapping the Five Laws of AI Governance to AI RMF 1.0, CSF 2.0 and SP 800-53 Rev 5.2.0, passed NIST screening and validation and were catalogued as Draft Informative References on 2026-10-06. They are in their 30-day public review. Catalog inclusion is an informative reference, not a NIST endorsement.
Where can I review the technical architecture? The full governance architecture — dispatch-time enforcement model, risk-tier classification, and framework control mappings — is publicly documented at Reference Documentation.
Filing record¶
Five provisional patent applications are on file with the United States Patent and Trademark Office, filed between April 4 and May 28, 2026.
United States provisional applications are not published by the USPTO. They remain unpublished unless and until a corresponding non-provisional application publishes, so these filings are not searchable in public patent databases today. Filing receipts are available to editors, counsel, and diligence reviewers on request.
What can be verified in the public record today: the three NIST OLIR Final informative references, the three Five Laws OLIR Draft references in public review, both Zenodo deposits and their DOIs, the ORCID record, and the AQ SCORE™ trademark filing (USPTO Serial 99850714).