AI systems

Systems that use automated models to make or assist decisions and to generate content across functions and settings; the interim response uses this practical, non-technical framing to identify where lifecycle measures (design, development, deployment, monitoring, decommissioning) apply.

Software that performs tasks using machine‑based inference and learning techniques; within the Plan this term covers systems deployed or considered for use by public administrations and subject to risk‑based assessment, documentation and alignment with responsible AI guidance.

Operational deployments composed of models, software, interfaces and integration components that perform tasks typically requiring human-like intelligence; distinguished from standalone trained models and including the system's runtime/data interactions and user-facing behavior.

Defined broadly to include machine learning models, knowledge‑based systems, natural language processing and computer vision systems used for decision‑support, automation or augmentation across sectors; the term links models and data to governance, accountability and human oversight obligations.

Software, models and algorithms that process data to produce recommendations or automated decisions for administrative tasks and public services; intended to cover algorithmic or computational systems deployed by public entities for service delivery and administrative activities.

Integrated technological solutions that implement AI technologies (algorithms, data processing, models, hardware) for specific practical applications across economic sectors, public services, or societal functions, delivering automated or augmented decision-making and task performance.

Systems of artificial intelligence are defined broadly in the Statute to encompass algorithmic systems and other software or automated decision‑making systems that process data to produce outputs or decisions; the term is used to capture the range of technologies subject to AESIA supervision and guidance.

Information-processing technologies that perform functions typically associated with human intelligence — including learning, reasoning, problem‑solving, perception, and language understanding — covering machine learning, deep learning, and other advanced computational techniques across the AI lifecycle.

The physical or virtual products or services that use AI to serve end users.

This term refers to a broad range of artificial intelligence technologies, from machine learning algorithms used for data analysis to more advanced applications like facial recognition. Their deployment requires careful ethical and legal scrutiny, as addressed by the toolkit.

Machine-based systems that can generate outputs such as predictions, recommendations, content, or other output influencing decisions made in real or virtual environments.

AI Systems encompass technologies that learn from data, make predictions, generate content, or operate autonomously. This broad category includes various forms of artificial intelligence that are the subject of the NIST SP 1326 framework.