EVE NeuroSystems LLC · USPTO Patent Portfolio
EVE AI Patent Portfolio
93 U.S. provisional patent applications across 6 coordinated families protecting EVE CoreGuard — deterministic AI governance infrastructure. Filed with the USPTO by EVE NeuroSystems LLC; invented by founder & creator Jamaurice Holt. U.S. Patent Pending.
The enforcement path
Three named anchors.
The spine of the portfolio traces a single request from authorization, to routing, to cryptographic proof — every AI action evaluated before it runs and signed for evidence after.
Authority Resolution
Deterministic rule evaluation that resolves whether a proposed AI action is authorized — in under a millisecond, before it can execute. The anchor of the Authority family, tracing to earliest-priority Serial No. 63/988,235.
Governed Inference Routing
Multi-axis model and provider selection under cost governance — routing every inference through policy rather than around it, so the choice of model is itself a governed, auditable decision.
Cryptographic Authority Chain
Ed25519-signed decision certificates in a hash-chained, tamper-evident trail — independently verifiable evidence for every governed decision, built for regulatory attestation.
Portfolio structure
Six coordinated families.
The 93 provisional applications are organized into six deliberately coordinated families spanning the full governance lifecycle — from pre-execution authority to cryptographic evidence, adversarial defense, and sovereign hardware identity — across the regulated domains of lending, healthcare, enterprise, and government.
| Patent family | Scope | Representative filing | Priority |
|---|---|---|---|
| Authority Resolution (Anchor) | Deterministic pre-execution authority resolution, cognitive veto, and pre-execution decision weighting — the power to refuse before any action runs. | 64/022,677 | Feb 22, 2026 |
| Governed Inference Routing (Routing) | Policy-governed model and provider selection with cost governance, unified policy-enforcement architecture, and control/execution/evidence planes. | 64/022,671 | Mar 20, 2026 |
| Cryptographic Authority Chain (Compliance) | Ed25519-signed decision certificates, hash-chained attestation, multi-framework regulatory scoring, and jurisdictional compliance. | 64/022,682 | Mar 23, 2026 |
| Evidence, Replay & Audit | Full-state governance decision replay, forensic reconstruction of decision timelines, cryptographic watermarking, and canary-token exfiltration detection. | 64/016,569 | Mar 25, 2026 |
| Adversarial Resilience & Threat Defense | Automated multi-model red-team loops, multi-turn privilege-escalation detection, enterprise AI circuit breakers, and hardware-enforced prompt-injection firewalling. | 64/014,751 | Mar 23, 2026 |
| Sovereign Identity & Hardware-Anchored Governance | Hardware-fused cognitive governance, multi-tenant digital sovereignty, hardware-anchored key derivation, and identity-drift budget management. | 63/988,324 | Feb 22, 2026 |
USPTO Patent Center
Provisional filing register.
Representative provisional applications across the portfolio, sourced directly from the USPTO Patent Center. All filings by Jamaurice Devron Holt on behalf of EVE NeuroSystems LLC, under 35 U.S.C. § 111(b).
| Application No. | Receipt | Filed | Provisional patent application |
|---|---|---|---|
| 64/016,528 | — | 03/25/2026 | Cryptographic Watermarking System for Governed AI Outputs with Independent Verification |
| 64/016,543 | — | 03/25/2026 | Canary Token Injection and Exfiltration Detection System for AI Governance Pipelines |
| 64/016,559 | — | 03/25/2026 | Information Density Budget Governance for AI Session Management |
| 64/016,569 | — | 03/25/2026 | Full-State Governance Decision Replay System for Regulatory Audit of AI Systems |
| 64/016,578 | — | 03/25/2026 | Policy-as-Code Engine for Deterministic AI Governance Compliance |
| 64/016,590 | — | 03/25/2026 | Recursive Governance Self-Verification Defense for AI Governance Systems |
| 64/014,751 | 74982908 | 03/23/2026 | System and Method for Adversarial Evolution of AI Governance Through Iterative Self-Hardening Using Automated Multi-Model Red Team Loops |
| 64/014,277 | 74976851 | 03/23/2026 | System and Method for Detecting and Preventing Multi-Turn Privilege Escalation in AI Governance Pipelines |
| 64/014,269 | 74977125 | 03/23/2026 | System and Method for Governance Isolation Preventing Philosophical Argument-Based Safety Dissolution in Governed AI Systems |
| 64/014,259 | 74977028 | 03/23/2026 | System and Method for Detecting Factual Contradictions in AI Outputs Using Verified Knowledge Base Comparison with Negation Matching |
| 64/013,866 | 74967810 | 03/23/2026 | System and Method for Regulatory Attestation of AI Decisions Through Multi-Framework Compliance Scoring and Hash-Chained Certificate Generation |
| 64/013,864 | 74967834 | 03/23/2026 | System and Method for Forensic Reconstruction of AI Governance Decision Timelines from Distributed Evidence Stores |
| 64/013,847 | 74967389 | 03/23/2026 | System and Method for Cross-Enterprise Federated AI Governance Through Signed Policy Hash Exchange and Trust-Scored Mesh Topology |
| 64/013,841 | 74967484 | 03/23/2026 | System and Method for AI Liability Assignment and Transfer Based on Governance Tier and Decision Verification State |
| 64/013,838 | 74967171 | 03/23/2026 | System and Method for Adversarial Resilience Scoring of AI Governance Systems Through Systematic Boundary Probing |
| 64/013,834 | 74967009 | 03/23/2026 | System and Method for Enterprise AI Circuit Breaker with Graduated State Recovery and Threshold-Based Automatic Triggering |
| 64/013,509 | 74963332 | 03/22/2026 | System and Method for Cryptographically Signed Self-Introspection in AI Systems with Claim Extraction and Automated Rewriting |
| 64/013,507 | 74963565 | 03/22/2026 | System and Method for Enforcing Governance State Continuity in AI Systems to Prevent Cold-Switching and Policy Disavowal |
| 64/013,503 | 74963544 | 03/22/2026 | System and Method for Computing Confidence-Reality Divergence in Artificial Intelligence Systems with Domain-Aware Threshold Calibration |
| 64/013,498 | 74963412 | 03/22/2026 | System and Method for Pre-Release Telemetry Verification of AI Outputs Using Multi-Stage Pipeline with Confidence-Reality Divergence Scoring |
| 64/013,493 | 74963425 | 03/22/2026 | System and Method for Organizing AI Governance into Control, Execution, and Evidence Planes with Deterministic Pre-Inference Policy Enforcement |
| 64/012,511 | 74959222 | 03/21/2026 | Recursive Autonomy Engine with Bounded Self-Improvement Through Outcome-Driven Pattern Analysis and Governed Proposal Generation for Artificial Intelligence Systems |
| 64/012,252 | 74957230 | 03/20/2026 | Proactive Cognitive Cost Assessment System with Capability Impairment and Persistent Residue for Artificial Intelligence Decision Integrity |
| 64/012,234 | 74957588 | 03/20/2026 | Protocol for Cross-Domain Sovereign Identity Handshaking and Non-Repudiable Agentic Negotiation |
| 64/012,224 | 74957747 | 03/20/2026 | Unified Policy-Enforcement Architecture for AI Systems with Deterministic Decision Resolution and Auditable Authority |
| 64/012,215 | 74957416 | 03/20/2026 | Pre-Execution Decision Weighting System Using Multi-Subsystem Policy Scoring for Governed Artificial Intelligence |
| 64/007,003 | 74874601 | 03/16/2026 | Autonomous Cognitive Executive Control Architecture Employing Deterministic Logical Gate Arrays and Proportional-Integral Control Loops for Pre-Response Orchestration in Autonomous Agent Systems |
| 64/006,976 | 74874210 | 03/16/2026 | Systems and Methods for Hardware-Enforced AI Security Employing Deterministic Logical Gate Arrays for Multi-Phase Prompt Injection Firewalling, Behavioral Fingerprinting with Proportional-Integral Controlled Threat Governance, and Byzantine Fault-Tolerant Insider Threat Detection on a PolarFire SoC Platform |
| 64/006,954 | 74873928 | 03/16/2026 | Epistemic Grounding & Social Intelligence |
| 63/999,256 | 74760077 | 03/07/2026 | Epistemic Refusal Apparatus and Data-Confidence-Gated Observability Methods for Anti-Marketing Health Reporting in Autonomous AI Systems |
| 63/999,252 | 74760522 | 03/07/2026 | Deterministic Hierarchical Key Derivation Apparatus and Action-Scoped Authentication Methods for Hardware-Anchored Autonomous Agent Governance |
| 63/999,243 | 74760253 | 03/07/2026 | Dynamic Governance Throttling System with Proportional Partial Application for Identity Drift Budget Management in Autonomous AI Agents |
| 63/999,238 | 74760427 | 03/07/2026 | Mathematically Falsifiable Invariant Certification System for Autonomous AI Financial Ledger Integrity and Cross-Domain Audit Verification Employing Canonical Decimal Quantization, Deterministic Hash Chain Linkage, and Hardware-Anchored Falsification Probes |
| 63/993,352 | 74664154 | 02/28/2026 | Context-Adaptive Governance Architecture with Pre-Authorization Envelopes: A System and Method for Amortized Stakes Classification, Neuro-Symbolic Constraint Translation, Mission-Scoped Governance Ticketing, Trust-Attenuated Delegation Chain Verification, and Federated Cross-System Governance Protocol for Autonomous Computational Agents |
| 63/992,157 | 74644921 | 02/27/2026 | Non-Deterministic Veto Architecture: A Hardware-Fused Cognitive Governance System for Invariant-Based Autonomous Integrity Preservation Against Agentic Hijacking and Adversarial Prompt Injection |
| 63/991,664 | 74637886 | 02/26/2026 | Deterministic Hardware-Software Interlock System for Autonomous Research Pipeline Orchestration, Recursive Self-Improvement Containment, Per-User Sovereign Guardianship, Tick-Based External Observation, Priority-State Vector Execution, and Tiered Autonomous Action Authorization |
| 63/991,649 | 74632794 | 02/26/2026 | Deterministic Hardware-Software Interlock System for Autonomous Research Pipeline Orchestration, Recursive Self-Improvement Containment, Per-User Sovereign Guardianship, Continuous External Observation, Goal-to-Action Execution Governance, and Tiered Autonomous Action Authorization |
| 63/989,479 | 74589333 | 02/24/2026 | Forensic Attribution, Energy-Adaptive Intelligence, and Jurisdictional Compliance Framework for Sovereign AI Agents |
| 63/989,260 | 74584486 | 02/24/2026 | Architectural Framework for Autonomous Agentic Sovereignty, Multi-Modal Provenance Verification, and Spatiotemporal Ethical Reasoning |
| 63/989,150 | 74583611 | 02/23/2026 | Systems for Causal Logic Synthesis, Autonomous Inventorship Detection, and Sovereign Identity Migration in Advanced AI Agents |
| 63/989,125 | 74583513 | 02/23/2026 | Governance-as-a-Service Framework for AI Integrity, Cryptographic Resilience Certification, and Belief Consistency |
| 63/989,113 | 74583387 | 02/23/2026 | Architectural Framework for AI Governance Continuity and Sovereign Policy Defense |
| 63/989,085 | 74582833 | 02/23/2026 | Advanced Cognitive Architecture for Autonomous Robotic Governance, Synthetic Intuition, and Existential Identity Defense |
| 63/988,235 | 74565934 | 02/22/2026 | Deterministic Hardware-Based Cognitive Veto System for AI Governance — Earliest priority |
| 63/988,324 (a) | 74566379 | 02/22/2026 | Deterministic Hardware-Fused Cognitive Governance System for Autonomous Agent Identity Preservation and Multi-Tenant Digital Sovereignty |
| 63/988,324 (b) | 74566604 | 02/22/2026 | Deterministic Hardware-Fused Cognitive Governance System for Autonomous Agent Identity Preservation and Multi-Tenant Digital Sovereignty |
| 63/988,291 | 74566425 | 02/22/2026 | AI Governance-as-a-Service and Digital Sovereignty Systems |
Representative filings from a portfolio of 93 U.S. provisional applications. Application 63/988,324 carries two receipt numbers (a & b). Data sourced directly from the USPTO Patent Center.
The architecture the portfolio protects
EVE CoreGuard is deterministic AI governance infrastructure: a layer that decides whether an AI action is allowed before the action executes, and produces cryptographic proof of that decision after. The portfolio's 93 provisional applications describe that layer end to end — from the moment a request arrives, through model selection, to the tamper-evident record left behind.
The design separates governance into distinct planes — control, execution, and evidence — with policy enforced deterministically at pre-inference time rather than inferred by the model itself. Authority resolution answers may this run in under a millisecond. Governed inference routing decides how it runs, selecting models and providers under explicit cost and policy constraints. The cryptographic authority chain then binds the outcome into an Ed25519-signed, hash-chained certificate that any party can verify independently — the foundation for regulatory attestation across lending, healthcare, enterprise, and government.
Around that spine sit the coordinated families that make governance survivable in the real world: full-state decision replay and forensic reconstruction for audit; adversarial red-team loops, privilege-escalation detection, and circuit breakers for resilience; and hardware-fused cognitive veto with multi-tenant sovereign identity for the cases where software assurances are not enough. Several filings extend the same discipline into hardware — deterministic logical gate arrays and PolarFire SoC platforms — so that the veto cannot be argued away.
The Infrastructure of No
Most AI safety work tries to make models say better things. EVE CoreGuard is built on the opposite conviction: real governance is the power to refuse. Its guiding principle is simple and absolute — no AI action executes without proof.
Governance is not a suggestion the model may follow. It is a gate the action must pass. If the authority cannot be resolved and the proof cannot be produced, the answer is no — deterministically, every time.
That is why the portfolio emphasizes deterministic evaluation over probabilistic judgment, cryptographic evidence over logs that can be edited, and hardware anchoring over policy that lives only in software. A refusal that can be talked out of is not governance; a decision that cannot be independently verified is not evidence. The Infrastructure of No is what remains when you insist on both — and it is what these filings, taken together, are designed to protect.
U.S. Patent Pending — a growing portfolio of provisional applications filed under 35 U.S.C. § 111(b), from earliest-priority Serial No. 63/988,235 onward. Assignee and filer: EVE NeuroSystems LLC. Inventor: Jamaurice Devron Holt.
Licensing, partnership, or a governance conversation.
The EVE AI portfolio is open to licensing discussions and collaboration on deterministic AI governance. Reach out to founder and creator Jamaurice Holt.
