Ternary Moral Logic: A Governance-Native Constitutional Architecture for Auditable Artificial Intelligence
Fuente:
Zenodo
Enregistré dans:
| Auteur principal: | |
|---|---|
| Format: | Recurso digital |
| Publié: |
Zenodo
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866901185152155648 |
|---|---|
| author | Goukassian, Lev |
| author_facet | Goukassian, Lev |
| contents | <p>The proliferation of high-risk Artificial Intelligence (AI) systems presents a critical "implementation gap" between normative ethical principles (e.g., UNESCO, OECD) and operational technical compliance. This monograph introduces <strong>Ternary Moral Logic (TML)</strong>, a constitutional architecture designed to bridge this gap by replacing binary decision logic with a triadic framework: <strong>+1 (Proceed), 0 (Sacred Zero/Pause), and –1 (Refuse)</strong>.</p> <p>The core innovation is the <strong>Sacred Zero</strong>, a mandatory state of hesitation triggered by uncertainty or ethical conflict, which enforces the "Goukassian Vow": <em>Pause when truth is uncertain; Refuse when harm is clear; Proceed where truth is</em>. This state activates <strong>Always Memory</strong> to capture a cryptographically sealed snapshot of the decision context, generating <strong>Moral Trace Logs (MTLs)</strong>.</p> <p>To resolve the "Operationalization Paradox" between safety and speed, TML employs a <strong>Dual-Lane Latency Architecture</strong>, separating ultra-low latency inference (<2ms) from asynchronous <strong>Merkle-Batched Anchoring</strong> (<500ms). This ensures that governance does not degrade performance in safety-critical domains (e.g., Autonomous Vehicles, Healthcare).</p> <p>This architecture provides the necessary technical controls to satisfy the risk management and quality assurance requirements of the <strong>EU AI Act (Articles 9 and 17)</strong>, <strong>NIST AI RMF</strong>, and <strong>ISO/IEC 42001</strong>. By leveraging <strong>Ephemeral Key Rotation (EKR)</strong> and decentralized anchoring, TML logs meet the authentication standards of <strong>FRE 901</strong> and <strong>eIDAS</strong>, rendering algorithmic decisions legally auditable and admissible as digital evidence.<br><br><strong><a href="https://fractonicmind.github.io/TernaryMoralLogic/Research_Reports/AI_Governance_Constitutional_Architecture_Console.html" rel="nofollow">Interactive Visualization</a>:</strong> A runtime simulator demonstrating the Dual-Lane Latency architecture and Real-Time Log generation is included with this monograph as AI_Governance_Constitutional_Architecture_Console.html</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_17849131 |
| institution | Zenodo |
| language | |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Ternary Moral Logic: A Governance-Native Constitutional Architecture for Auditable Artificial Intelligence Goukassian, Lev Ternary Moral Logic AI Governance Sacred Zero Auditable AI EU AI Act Compliance Merkle-Batched Anchoring AI Ethics Moral Trace Logs Goukassian Vow Algorithmic Accountability Digital Evidence ISO 42001 Safety-Critical Systems <p>The proliferation of high-risk Artificial Intelligence (AI) systems presents a critical "implementation gap" between normative ethical principles (e.g., UNESCO, OECD) and operational technical compliance. This monograph introduces <strong>Ternary Moral Logic (TML)</strong>, a constitutional architecture designed to bridge this gap by replacing binary decision logic with a triadic framework: <strong>+1 (Proceed), 0 (Sacred Zero/Pause), and –1 (Refuse)</strong>.</p> <p>The core innovation is the <strong>Sacred Zero</strong>, a mandatory state of hesitation triggered by uncertainty or ethical conflict, which enforces the "Goukassian Vow": <em>Pause when truth is uncertain; Refuse when harm is clear; Proceed where truth is</em>. This state activates <strong>Always Memory</strong> to capture a cryptographically sealed snapshot of the decision context, generating <strong>Moral Trace Logs (MTLs)</strong>.</p> <p>To resolve the "Operationalization Paradox" between safety and speed, TML employs a <strong>Dual-Lane Latency Architecture</strong>, separating ultra-low latency inference (<2ms) from asynchronous <strong>Merkle-Batched Anchoring</strong> (<500ms). This ensures that governance does not degrade performance in safety-critical domains (e.g., Autonomous Vehicles, Healthcare).</p> <p>This architecture provides the necessary technical controls to satisfy the risk management and quality assurance requirements of the <strong>EU AI Act (Articles 9 and 17)</strong>, <strong>NIST AI RMF</strong>, and <strong>ISO/IEC 42001</strong>. By leveraging <strong>Ephemeral Key Rotation (EKR)</strong> and decentralized anchoring, TML logs meet the authentication standards of <strong>FRE 901</strong> and <strong>eIDAS</strong>, rendering algorithmic decisions legally auditable and admissible as digital evidence.<br><br><strong><a href="https://fractonicmind.github.io/TernaryMoralLogic/Research_Reports/AI_Governance_Constitutional_Architecture_Console.html" rel="nofollow">Interactive Visualization</a>:</strong> A runtime simulator demonstrating the Dual-Lane Latency architecture and Real-Time Log generation is included with this monograph as AI_Governance_Constitutional_Architecture_Console.html</p> |
| title | Ternary Moral Logic: A Governance-Native Constitutional Architecture for Auditable Artificial Intelligence |
| topic | Ternary Moral Logic AI Governance Sacred Zero Auditable AI EU AI Act Compliance Merkle-Batched Anchoring AI Ethics Moral Trace Logs Goukassian Vow Algorithmic Accountability Digital Evidence ISO 42001 Safety-Critical Systems |
| url | https://doi.org/10.5281/zenodo.17849131 |