UT Libraries Metadata Policy

v.1, 04/09/26, Devon Murphy, Jennifer Bussey, Mirko Hanke, Katie Pierce Meyer, Sean O’Bryan, Ryan Sullivant

Purpose

This policy establishes institutional standards for metadata creation, management, and governance to ensure discoverability, preservation, interoperability, and equitable access to library and archival collections.

Definition

Metadata is functional data that provides administrative context, description, and structural information for the University of Texas Libraries' (UT Libraries) collections. Metadata provides crucial organizational structure for both physical and digital collections, supporting overall management, discovery, access, preservation, and compliance within and beyond our institution.

Guiding Principles

Metadata creation follows professional standards and ethics that ensure accurate and respectful representation of assets, agents, and subjects.
Metadata processes balance professional standards through a “good enough” approach that prioritizes consistent access to library assets over descriptive perfection.
Metadata is interoperable across systems, privileging broader schemas and structures over the bespoke.
Metadata processes and workflows are consistently documented to engender transparency and accountability.
Metadata improves accessibility, long-term sustainability, and preservation of assets through standards compliance.

Scope

This policy applies to all metadata created, managed, or aggregated by UT Libraries across all platforms, systems, and collections in all library units and locations, including but not limited to:

AI-generated metadata.
MARC and non-MARC metadata.
Vendor-supplied and partner-contributed metadata.

Governance

Metadata policy, guidelines, and best practices shall be created and managed by the Metadata Steering Committee and the Metadata Analyst. The Metadata Analyst’s position carries out implementation of these decisions.

Content Standards

Metadata practices shall be coordinated across organizational units to ensure consistency, reduce duplication, and promote interoperability. This includes standardization of controlled vocabularies and authorities, metadata fields and schemas, and implementation of best practices for harmful content and language. Local variation is allowable when relevant, such as the use of local vocabularies in post-custodial collections. Significant changes shall be discussed with the Metadata Analyst and Metadata Steering Committee to allow for documentation and crosswalking.

Interoperability

Systems used by UT Libraries shall support open technical standards and enable metadata portability to prevent vendor lock-in and ensure long-term sustainability of assets. In turn, systems shall support metadata reuse workflows to reduce duplication of information and synchronize similar assets across the library resource ecosystem.

Maintenance

Metadata shall be reviewed by relevant staff, (e.g. collection curators, catalogers), and the Metadata Analyst, on a consistent schedule in accordance with shared quality measures, ensuring accuracy and consistency with institutional and professional best practices. Items maintained include:

Schemas.
Controlled vocabularies and authorities.
Local descriptive practice.

Metadata shall be created and maintained with the following quality measures:

Accuracy: metadata correctly represents the asset.
Completeness: metadata fully describes the asset; all required elements are present.
Consistency: metadata follows established standards and local practices.
Currency: metadata is updated to match evolving standards.
Ethics: metadata is made accessible and representative.

All metadata practices must be documented in a guideline or best practice format, and shared to the UT Libraries Metadata Documentation Wiki.

Artificial Intelligence

Metadata created or modified by artificial intelligence and related automated tools shall:

Be evaluated for bias, accuracy, and appropriateness.
Be clearly labeled as created or modified by AI, with AI usage and workflows made transparent.
Be compliant with institutional policies on AI use and data privacy.

Copyright

Metadata created by UT Libraries is not copyrightable and is open to be shared, with exceptions for privacy, security, contractual obligations, or graphic/sensitive content. Vendor-supplied metadata shall be negotiated to maximize reuse. Metadata sharing agreements with external partners, such as other institutions or vendors, must use appropriate institutional procedures such as memorandums of understanding (MOUs).

Metadata Policy