cycle of life
Stages from R&D through deployment and maintenance.
How the laws define this
Both regulatory instruments establish that an AI system's cycle of life is an end-to-end framework extending from initial creation through active deployment to final decommissioning, ensuring legal obligations apply across every phase of existence. Where these provisions diverge is in the specificity of the lifecycle stages and the particular regulatory duties attached to them. Expediente 1472-D-2023 – Modificaciones a la Ley 25467 (Ciencia, Tecnología e Innovación) relativas a Inteligencia Artificial (Amendment to Science & Tech law regarding AI) (bill) categorizes this progression into broader phases—research and development, deployment, operational use, maintenance, and decommissioning—and uses them to determine the scope of registration and review requirements. Conversely, Resolução CNJ nº 615/2025 details a highly granular progression explicitly listing conception, design, data curation, development, training, testing, deployment, operation, monitoring, maintenance, versioning, and decommissioning, serving as the framework for a wider set of governance, documentation, retention, and audit duties.
Synthesised from the 2 statutory definitions below. Each one is quoted in full, with its source.
How jurisdictions differ
What is distinctive in each definition, binding law in force first.
| Jurisdiction | Instrument | Status | What is distinctive |
|---|---|---|---|
| Brazil | CNJ Resolution No. 615/2025 | In Force | Adds data curation, training, testing, versioning and decommissioning to the stages, and explicitly uses it as a framework for governance, documentation, retention and audit obligations. |
| Argentina | Bill 1472-D-2023 – Amendments to Law 25467 (Science, Technology and Innovation) regarding Artificial Intelligence | Withdrawn | Adds research and operational use to the stages, and specifies that registration and review requirements apply to AI systems at any stage. |
What it triggers
Definitions (2)
The lifecycle concept used to determine the scope of obligations under the amendment, encompassing stages of an AI system from research and development, through deployment and operational use, to ongoing maintenance and decommissioning; registration and review requirements apply to AI systems at any stage of this cycle.
The end‑to‑end lifecycle concept covering conception, design, data curation, development, training, testing, deployment, operation, monitoring, maintenance, versioning and decommissioning of AI systems, used as the framework for applicable governance, documentation, retention and audit obligations.
Not to be confused with
Lifecycle covers all stages from design through decommissioning, while cycle of life definitions specify additional stages like data curation or research and link to obligations.
Activities within the lifecycle refers to stages from design to decommissioning, whereas cycle of life definitions include more stages like training or operational use and tie to duties.
AI lifecycle phases lists only design, training, deployment, and monitoring, narrower than cycle of life definitions that add testing, operation, maintenance and decommissioning.
Questions and answers
What obligations does the cycle of life trigger in Brazilian judiciary AI rules?
CNJ Resolution No. 615/2025 uses the cycle of life as the framework for governance, documentation, retention and audit obligations for AI systems.
Does the cycle of life in Argentina's bill apply to all AI stages?
Bill 1472-D-2023 states that registration and review requirements apply to AI systems at any stage of the cycle of life, from research to decommissioning.
How does cycle of life differ from basic lifecycle in AI regs?
Cycle of life definitions expand on stages like adding data curation or research, and explicitly link to compliance duties unlike simpler lifecycle descriptions.
Related Terms
Lifecycle
All stages from design through deployment, operation, and decommissioning....
Cycle of Life (Ciclo de Vida)
End‑to‑end lifespan of an AI system from design to decommission....
activities within the lifecycle
Stages of an AI system from design to decommissioning....
AI lifecycle phases
Design, training, deployment, and monitoring stages of AI systems....
AI life cycle
Stages from scoping through decommissioning....