Collections Policy

Museum TN collects and exhibits evidence that helps visitors understand the history, workings, social impact and possible futures of Artificial Intelligence.

What counts as an exhibit

An exhibit may be a photographed physical object, historical document, scientific paper, source code, software, interface, dataset or model card, institutional record, or another born-digital object. Physicality is not required; provenance, documentation and relevance are.

What is not presented as historical evidence

AI-generated imagery, artistic renderings, diagrams and curatorial reconstructions are never presented as originals. When used, their role is explicitly labelled.

Acquisition and selection

Material enters the collection when it answers a clear curatorial question, has adequate sourcing and can be presented lawfully. Primary sources, institutional archives, open licences and born-digital evidence are preferred.

Rights

Being available online is not permission to reuse an image. The museum uses its own material, sufficiently clear public-domain material, Creative Commons material under its terms, or material used with permission. Where visual rights are unclear, the museum uses a metadata-based documentary facsimile and links to the primary source instead.

Documentation

Every standalone exhibit has a collection number, date, evidence type, primary source, visual-representation statement, image-rights statement and curatorial-review date. Relationships between exhibits are also recorded as part of the collection.

Correction and deaccession

The collection is living. Evidence or representations may be corrected, relabelled or withdrawn when stronger documentation, rights issues or better curatorial solutions emerge. Established exhibit URLs are preserved where possible.

Access

The collection is free and requires no account. Documentation is public-facing rather than hidden as internal museum paperwork.