GovernanceCore
Foundations

AI system

Working definition

Reviewed 30 July 2026

An AI system is a machine-based system that, for explicit or implicit objectives, infers from its inputs how to generate outputs such as predictions, content, recommendations, or decisions. Systems vary in autonomy and in how adaptive they are after deployment.

Context

Why it matters

Governance decisions made only at model level can miss risks introduced by workflow design, human reliance, integrations, data, and downstream use.

Operating note

What this looks like in practice

  1. 01Document the whole sociotechnical system, not only the algorithm.
  2. 02Record intended purpose, users, affected people, environment, and decision pathway.
  3. 03Reassess the system when its context, autonomy, model, or data materially changes.

Sources & further research

Primary authority anchors the definition. Research links add conceptual or operational depth. External sources may update independently; always verify legal duties against the current official text.

  1. Official source

    Explanatory memorandum on the updated OECD definition of an AI system

    Explains the internationally influential 2023 revision of the OECD definition.

    OECD.AI
    2024    Open source ↗
  2. Official source

    Regulation (EU) 2024/1689, Article 3(1)

    The EU AI Act definition closely tracks the OECD formulation and is legally operative in the EU.

    European Union
    2024    Open source ↗