Glossary

Mechanisms that determine who can access which data, systems, or actions.

AI conformity assessment is the process of demonstrating that an AI system meets required legal and technical requirements before it is placed on the market or put into use.

AI documentation requirements are the records and technical documents needed to explain an AI system’s purpose, design, testing, controls, and limitations for compliance and safe use.

Algorithmic transparency is providing clear information about how an AI system works, what data it uses, and what its limits and failure modes are.

A tamper-evident record of actions, decisions, and system events over time.

Compliance-aware retrieval applies governance and regulatory constraints (access, provenance, logging) so systems retrieve and cite sources in a defensible way.

The policies, roles, and processes that ensure data is managed responsibly and compliantly.

Where data is physically stored or processed, often constrained by law or policy.

Rules defining how long different types of data are kept and when they must be deleted.

The EU AI Act is the European Union’s risk-based regulation for artificial intelligence, setting obligations for providers and deployers depending on system risk.

A high-risk AI system is an AI system classified as high-risk under the EU AI Act due to its intended use, triggering stricter obligations and controls.

Human oversight means people can understand, monitor, and intervene in an AI system’s operation, including the ability to override or stop it.

The extent to which humans can understand how a model makes its predictions.

An approach where privacy and data protection are built into systems from the outset.

Regulatory drift detection monitors changes in laws and guidance that can make an AI system’s retrieval, outputs, or controls outdated.

An access control model where permissions are assigned to roles, not directly to individual users.

Information about where data or content came from and how it was produced.