Global Sigma Public Standard: Open framework for certification of high-risk AI systems
Global Sigma Public Standard: Open framework for certification of high-risk AI systems
GLOBAL SIGMA Executive Summary / Resumen Ejecutivo ENGLISH 🔍 THE CHALLENGE Legal frameworks (EU AI Act) mandate WHAT to do. Management standards (ISO 42001) guide HOW to govern. BUT neither provides VERIFIABLE, AUDITABLE properties for High-Risk AI Systems. ✅ OUR RESPONSE Global Sigma operationalizes 12 Principles (P1-P12) from: EU AI Act (Articles 9-15) NIST AI RMF OECD AI Principles UNESCO Ethics Recommendations ISO/IEC 23894 💡 KEY INNOVATION Two distinct paths: Technical Conformity Certification P1-P12 adherence Conformity + immutability No immutability required Cryptographic/forensic proof required No formal auditor Independent third-party audit Suitable for development MANDATORY for Tier 3/4 in regulated jurisdictions ⚠️ VETO LOGIC A system CANNOT be certified if it has: Veto Condition Type D2 P2 <80% reproducibility Irremediable D4 P8 <85% escalation Irremediable D11 Modifiable logs without trail Irremediable - P7 Non-compliant All vetoes default to irremediable 📊 RISK TIERS Tier Classification Certification 1 Minimal Risk Not Applicable 2 Limited Risk Voluntary 3 High Risk MANDATORY* 4 Critical Risk MANDATORY (enhanced) If regulated jurisdiction applies ESPAÑOL 🔍 EL DESAFÍO Los marcos legales (EU AI Act) exigen QUÉ hacer. Los estándares de gestión (ISO 42001) guían CÓMO gobernar. PERO ninguno proporciona PROPIEDADES VERIFICABLES Y AUDITABLES para Sistemas de IA de Alto Riesgo. ✅ NUESTRA RESPUESTA Global Sigma operacionaliza 12 Principios (P1-P12) de: EU AI Act (Artículos 9-15) NIST AI RMF Principios de IA de la OCDE Recomendaciones de Ética de UNESCO ISO/IEC 23894 💡 INNOVACIÓN CLAVE Dos vías distintas: Conformidad Técnica Certificación Adherencia a P1-P12 Conformidad + inmutabilidad Sin inmutabilidad requerida Prueba criptográfica/forense requerida Sin auditor formal Auditoría independiente de tercera parte Adecuado para desarrollo OBLIGATORIO para Tier 3/4 en jurisdicciones reguladas ⚠️ LÓGICA DE VETO Un sistema NO PUEDE ser certificado si tiene: Veto Condición Tipo D2 P2 <80% de reproducibilidad Irremediable D4 P8 <85% de escalamiento Irremediable D11 Logs modificables sin rastro Irremediable - P7 No conforme Todos los vetos se vuelven irremediables 📊 NIVELES DE RIESGO Nivel Clasificación Certificación 1 Riesgo Mínimo No Aplica 2 Riesgo Limitado Voluntaria 3 Alto Riesgo OBLIGATORIA* 4 Riesgo Crítico OBLIGATORIA (reforzada) Si aplica jurisdicción regulada GLOBAL SIGMA PUBLIC STANDARD - Research Validation Edition (2026-2027) - Autor: Dra. Paula Melissa Alfaro Montero - Licencia: CC BY-SA 4.0 - Repositorio: github.com/lunalfaro/global-sigma-standard-
Primary: Computer Science - Artificial Intelligence Secondary: Law and Technology (Regulatory Compliance) Technical: Information Systems Audit and Forensic Verification Ethical: Human Rights and Algorithmic Fairness Engineering: Software Modularity and Resilience, Primary: Computer Science - Artificial Intelligence Secondary: Law and Technology (Regulatory Compliance) Technical: Information Systems Audit and Forensic Verification Ethical: Human Rights and Algorithmic Fairness Engineering: Software Modularity and Resilience, AI Governance, AI Certification, High-Risk AI Systems, EU AI Act Compliance, Technical Conformity, Reproducible Determinism, Algorithmic Audit, Immutability Patterns, AI Ethics, Trustworthy AI, Risk Tier Classification, Open Framework, Knowledge Sovereignty, Adversarial Security, Continuous Auditability, Standard for AI.
Primary: Computer Science - Artificial Intelligence Secondary: Law and Technology (Regulatory Compliance) Technical: Information Systems Audit and Forensic Verification Ethical: Human Rights and Algorithmic Fairness Engineering: Software Modularity and Resilience, Primary: Computer Science - Artificial Intelligence Secondary: Law and Technology (Regulatory Compliance) Technical: Information Systems Audit and Forensic Verification Ethical: Human Rights and Algorithmic Fairness Engineering: Software Modularity and Resilience, AI Governance, AI Certification, High-Risk AI Systems, EU AI Act Compliance, Technical Conformity, Reproducible Determinism, Algorithmic Audit, Immutability Patterns, AI Ethics, Trustworthy AI, Risk Tier Classification, Open Framework, Knowledge Sovereignty, Adversarial Security, Continuous Auditability, Standard for AI.
selected citations These citations are derived from selected sources.This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).0 popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.Average influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).Average impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.Average
Citations provided by BIP!Popularity provided by BIP!