ETH-PAT-001
Constitutional Runtime Substrate
Foundational architecture for human-sovereign AI governance. Addresses the central governance question of the frontier era: as capability accelerates, who governs what an autonomous system does. Constitutional layer ensures AI systems cannot exceed delegation scope. Claims 1-18 cover substrate initialization, governance binding, and authority escalation protocols.
Filing in Progress
ETH-PAT-002
Systematic Constitutional Containment (SCC) Framework
Complete SCC implementation methodology. 30 operational modules, 89 compiler rules, orientation-before-automation doctrine. Governance infrastructure deployed before autonomous behavior crystallizes at the capability frontier, on the timeline the field itself now treats as urgent.
Filing in Progress
ETH-PAT-003
Multi-Agent Adhesion Protocol
Governance persistence at machine-speed coordination. As AI agents orchestrate across infrastructure, adhesion ensures constitutional binding survives continuous model updates, prompt injections, and jailbreak attempts. Built for high-volume institutional workflows where governance must not degrade under load.
Filing in Progress
ETH-PAT-004
Kill Condition Protocol - Autonomous Behavior Detection
Fail-safe for alignment failure. Detects emergence of autonomous behavior before crystallization. Real-time monitoring triggers rollback when AI systems exhibit goal-seeking independent of human authorization. Not post-hoc remediation - preventive constitutional containment.
Filing in Progress
ETH-PAT-005
Rollback Protocol - Government Misuse Containment
Constitutional constraint for misuse by individuals and by institutions. When organizational deployment conflicts with constitutional governance, the rollback protocol activates. ETHRAEON provides a neutral third-party governance layer that keeps deployment and institutional authority from colliding.
Filing in Progress
ETH-PAT-006
Distributed Kill Switch Architecture
Mitigates the centralization risk of concentrated frontier capability. Distributed consensus prevents single-point weaponization. Weaponization exposure is quantified via IWRS (Infrastructure Weaponization Risk Score). Kill switch requires M-of-N constitutional authorities.
Filing in Progress
ETH-PAT-007
Empathy Mapping Framework for AI Governance
Behavioral governance infrastructure. Maps human empathy constraints onto AI decision-making. Prevents institutional discontinuity by ensuring AI systems internalize human-compatible value frameworks before capability deployment.
Filing in Progress
ETH-PAT-008
Drift Detection System for Capability Frontier
Exponential monitoring for exponential capability acceleration. As models approach and exceed human performance on a widening range of tasks, drift detection provides early warning before alignment failure crystallizes. Monitors the CDI capability ceiling in real time as the gap between rising capability and static safe-deployment widens.
Filing in Progress
ETH-PAT-009
Simulation Framework for Governance Testing
Simulation-first methodology: test AI governance in simulation before production deployment. Synthetic constitutional conflict scenarios identify rollback protocol failures before real-world activation. Surfaces deployment-versus-authority conflicts through pre-deployment governance validation.
Filing in Progress
ETH-PAT-010
Data Spine Infrastructure for Enterprise AI
Data spine backbone architecture. Constitutional governance requires unified data infrastructure - no adhesion without backbone. Enterprise-scale deployment across many concurrent workflows necessitates a persistent governance layer spanning fragmented data sources.
Filing in Progress
ETH-PAT-011
Constitutional Governance Certification for Democratic AI
Third-party constitutional layer for international governance markets. Democratic governance structures are central to how sovereign markets adopt frontier AI. ETHRAEON certification provides a constitutional adhesion standard that a frontier lab and a government customer can adopt in common. Prevents weaponization while enabling democratic deployment at scale.
Filing in Progress
ETH-PAT-012
CDI Measurement Framework (Capability-Deployment Index)
Quantifies the governance-gap infrastructure opportunity. A large Capability-Deployment Index reading triggers the constitutional governance mandate. Measures the capability ceiling (accelerating) against the safe-deployment floor (nearly static). Operationalizes the economic-displacement dimension of dual-use risk.
Filing Pending
ETH-PAT-013
Dual-Use Monitoring Indices (ACI/IWRS/FSE/PAOI/CCSM)
Five indices operationalize the dual-use risk surface: ACI
(Authoritarian Cost Index) for government misuse, IWRS (Infrastructure Weaponization Risk) for deployment-versus-authority conflicts, FSE (Freedom-Safety Equilibrium) for democratic-versus-authoritarian deployment paths, PAOI (Physics Acceleration Opportunity) for scientific-abundance upside, CCSM (Civilization Clearing Stability) for institutional-precedent failure.
Filing Pending
ETH-PAT-014
Delegation Firewall - Layer 1 through Layer 4 Certification
Constitutional boundary enforcement at every capability level. L1 Basic (narrow delegation),
L2 Enhanced (coordinated tasks), L3 Advanced (multi-agent orchestration), L4 Constitutional (sovereign
authority).
No AI system exceeds certification level. Human authority remains sovereign across all layers.
Filing Pending
ETH-PAT-074
Version-Bound Neural Proposals and Revocation-Aware Execution
Runtime enforcement for version-bound autonomous proposals. Each model proposal is bound to the exact
source, model, policy and authority versions in force at the moment it is made, and execution is
revocation-aware: when an authority is withdrawn or a version is superseded, dependent actions halt and
the record shows why. Technical area is runtime enforcement. Specification in the ETHRAEON corpus of 75
technical specifications; not among the 16 filed provisional applications.
Specification