Canberra's Duplicate Image Problem: The Key Decisions That Will Shape What Comes Next

A quiet but costly administrative problem is building inside Canberra's public sector. Duplicate images — identical or near-identical digital files stored multiple times across agency records systems — have accumulated across ACT and federal government departments to the point where storage costs, retrieval errors and compliance risks are now forcing a reckoning. The immediate question is not whether to act, but which approach to take and who pays for it.

The issue has sharpened because of two converging pressures. The ACT Government's Digital Strategy, which runs through to 2026, set agency-level targets for records consolidation, and several departments are approaching the end-of-financial-year audit period having not met them. At the same time, the federal government's data centre consolidation program — which affects agencies clustered along Northbourne Avenue and in Barton — has exposed just how many redundant files have been quietly accumulating since the shift to cloud storage accelerated around 2019.

Where the Problem Is Concentrated

The duplication issue is particularly acute in agencies that handle large volumes of visual evidence, identity documents or spatial data. The Australian Federal Police, headquartered on Majura Avenue in Majura Park, manages case-file imagery that can run to thousands of photos per investigation. The Australian Bureau of Statistics, based in Belconnen, maintains georeferenced image datasets that are updated cyclically — and older versions are rarely purged. Neither agency has confirmed the scale of its internal duplication backlog, and neither has been approached for comment for this story.

At the Australian National University in Acton, the University Library's digital collections team has been working since early 2025 on a deduplication project covering its digitised photograph holdings, which include material from the Noel Butlin Archives Centre. The work is funded partly through the National Collaborative Research Infrastructure Strategy program. A separate but related effort is underway at the National Archives of Australia, whose main reading room and digital access services are located in Mitchell, where staff have flagged in internal planning documents that legacy scanning workflows produced significant file duplication during the 2015–2020 digitisation push.

University of Canberra researchers based at the Bruce campus have published work on hash-based deduplication methods — a technical approach that compares unique file signatures rather than reading image content — as a cost-effective option for government-scale archives. The method can cut storage requirements substantially, though precise figures vary widely depending on the composition of a given dataset.

The Decisions That Now Have to Be Made

Three choices will define how this plays out over the next 12 months. First, agencies must decide whether to run deduplication internally using existing IT staff or procure specialist contractors — a decision with real budget consequences, given that Canberra-based digital records contractors currently quote day rates that can exceed $1,500 for senior archival data specialists. Second, institutions need to agree on retention rules before any file is deleted: a duplicate that looks redundant may be the only copy with intact metadata, which has legal significance under the Archives Act 1983. Third, the ACT and federal governments need to decide whether to coordinate their efforts or let each agency proceed independently, which risks duplication of the deduplication effort itself — a bureaucratic irony that is not lost on people working in this space.

The ACT Government's Chief Digital Officer directorate, which operates out of offices in Canberra City, is understood to be developing guidance for territory agencies, though no public consultation has been announced. The National Archives released updated digital preservation guidance in March 2026, but it stops short of mandating specific deduplication timelines.

For Canberrans who interact with these systems — whether lodging Freedom of Information requests, accessing digitised heritage collections at the Noel Butlin Archives, or relying on AFP case records — the practical stakes are real. Duplicate files slow retrieval systems, inflate FOI processing times, and can cause version-control errors when agencies share imagery across platforms. Getting the governance framework right before the deletion starts is the only way to avoid creating new problems while solving old ones.