Post-Market Monitoring Procedure

Purpose

This procedure establishes a comprehensive, systematic, and proportionate post-market monitoring system for high-risk AI systems to actively and continuously collect, document, and analyze data on the performance of AI systems throughout their operational lifecycle, in compliance with EU AI Act Article 72. The post-market monitoring system ensures: - Continuous verification of AI system compliance with EU AI Act requirements - Early detection of performance degradation, bias drift, and safety issues - Proactive identification of reasonably foreseeable risks not identified during pre-deployment assessment - Integration with risk management, quality management, and incident reporting systems - Evidence-based continuous improvement of AI systems

Applies To

• All high-risk AI systems placed on the market or put into service (per Annex III classification)
• All AI systems subject to conformity assessment under Article 43
• All AI systems deployed in production environments serving end users
• All AI systems processing personal data or making decisions affecting fundamental rights
• All AI systems that have completed deployment phase (post Phase 5 in lifecycle)
• AI systems in development or testing phases (covered by AI Development Lifecycle Procedure (PROC-AI-DEV-001))
• AI systems not yet deployed to production
• Minimal-risk AI systems (unless organization voluntarily applies monitoring)
• AI systems operated by deployers only (deployer obligations under Article 26)
• Research and development prototypes not intended for market placement