Search Museum Next

Unlocking Hidden Stories: How AI is Transforming Audiovisual Collections at the NFSA

Searching inside audiovisual material can reveal stories that catalogue descriptions miss, but it can also expose sensitive content through new routes. At the National Film and Sound Archive of Australia (NFSA), Jacqui Uhlmann, Head of Collection, and Tamara Osicka, Curatorial Operations Assistant Manager, described how an AI pilot brought discovery and access controls into the same conversation.

Searching inside audiovisual material

The NFSA’s vast audiovisual collection spans decades of Australian history, comprising everything from feature films and radio broadcasts to oral histories and video games. Traditionally, curators and archivists have relied on catalogue records and institutional expertise to navigate this treasure trove. However, as the collection continues to grow, these traditional methods face significant limitations.

The pilot explored facial and object recognition, speech-to-text transcription, and optical character recognition. Uhlmann and Osicka reported that these technologies improved the NFSA’s ability to search and interpret its collection, surfacing previously hidden or overlooked material. They described searches that could return relevant material in seconds where a curator might have spent hours searching manually.

A related approach to archival discovery can be found in IWM’s sound and image archive work.

When discovery reaches sensitive content

While AI offers incredible efficiency and insight, it also brings a host of ethical and practical challenges. Sensitive content, including culturally significant Indigenous materials and explicit audiovisual content, must be handled with care. AI’s capacity for rapid and sweeping analysis raises important questions about how to prevent inadvertent exposure to distressing material.

During the AI pilot, the NFSA team discovered that traditional checks and balances embedded in catalogue records could sometimes be bypassed by AI searches. This required them to rethink access controls and ensure that both staff and systems were prepared to handle potentially sensitive discoveries.

The pilot exposed a connection between discovery and access. Searching within audiovisual material could reveal content that catalogue-based safeguards had previously helped manage. For the NFSA, improving search therefore also meant reconsidering how sensitive results reached people, bringing collection knowledge and access decisions into the design of the system.

Lessons Learned from the AI Pilot

The pilot brought technical teams, curators and collection managers into a shared process. Their different skills and collection knowledge helped shape both implementation and the assessment of results. The account emphasised that AI could assist with searching, while contextual understanding, ethical decisions and nuanced judgement remained human responsibilities.

Sensitive material made these responsibilities concrete. Content warnings, access controls and staff training were among the measures the team identified as important. The work also called for ongoing learning, feedback and iterative improvements as the tools and staff experience developed.

Practical questions remained about cost, computational resources and maintaining a system at collection scale. Large result sets required attention to search, indexing and interfaces so people could navigate them. The team also highlighted data sovereignty: where material is stored, how it is processed and who can access it.

What the pilot left to resolve

The NFSA’s pilot programme raised further ambitions for audiovisual collection management. The team envisaged AI tools that could:

  • Generate accurate transcripts and summaries.
  • Identify errors in catalogue records.
  • Flag sensitive or harmful content more effectively.
  • Handle complex searches across both catalogue data and audiovisual content.

A Cautious Optimism

As Jacqui Uhlmann and Tamara Osicka emphasised, AI is not a silver bullet—it’s a tool, one that requires careful oversight, ongoing learning, and an institutional commitment to ethical practice. The NFSA’s journey showcases not just the potential of AI, but also the importance of balancing innovation with responsibility.

For cultural institutions worldwide, the NFSA’s experience offers a valuable blueprint for integrating AI into collection management—a journey filled with discovery, reflection, and transformation.

–

Learn more about how museums are using AI to revolutionise collections at MuseumNext’s Museum AI Summit.

About the author – Manuel Charr

Manuel Charr is a journalist working in the arts and cultural sectors. With a background in marketing, Manuel is drawn to arts organizations which are prepared to try inventive ways to reach new audiences.

Related Content

Turning Collections Into Data: Unlocking Hidden Stories at the Science Museum Group

What happens when museums stop treating their collections as static objects, and instead see them as data? That’s the question Daniel Belteki, Digital Research Applications...

Unlocking Curiosity: Transforming the Rijksmuseum with an Escape Room Experience

What happens when you combine the intrigue of a secret society with the treasures of one of the world’s most famous museums? This was the...

What a Dementia Programme Can Teach Us About the Hidden Power of Museums

What do railways, forests, Eurovision, and old suitcases have in common? On the surface, not much. But for people living with dementia in Liverpool—and increasingly...

Subscribe to the latest museum thinking

Fresh ideas from museums around the globe in your inbox each week