Lifecycle

The entire duration of an AI system's existence, including design, development, use, and decommissioning.

The sequence of phases through which an AI system passes—design, development, deployment, monitoring, and decommissioning—used to orient governance, risk management, and control activities across the system's operational life.

The sequence of phases covering conception, design, development, testing, deployment, monitoring, maintenance and retirement of an AI system. The lifecycle framing mandates that principles, controls, and assessments be applied at each stage to ensure continuous compliance and risk mitigation.

All stages of an AI system including design, research and development, testing, deployment, operation, maintenance and monitoring, through to decommissioning; the Convention requires iterative risk and impact assessment processes across these stages.

The defined set of stages applied to AI systems for governance and oversight: plan & design; data collection and processing; model building/adaptation; testing/validation; deployment; operation/monitoring; and decommissioning; each stage requires specific documentation, risk assessment and control measures under the strategy.