Canberra's Digital Archives Are Riddled With Duplicate Images — and the Numbers Are Worse Than Anyone Admitted

Roughly one in every three images stored across ACT government digital asset systems is a duplicate, according to figures circulated among ICT officers within the directorate network this financial year. The finding has triggered a quiet but urgent push to automate duplicate detection across agencies, with at least two major platforms already flagging the problem as a drag on performance and procurement costs.

The issue lands at a sensitive moment. The ACT government is mid-cycle on its whole-of-government digital infrastructure review, and agencies are under pressure to justify cloud storage expenditure ahead of the 2026-27 budget. Duplicate image files — the same photograph or graphic stored under different filenames, in different folders, or re-uploaded across multiple content management systems — inflate storage tallies without adding any functional value. At scale, across an organisation like the ACT Public Service, which employs roughly 23,000 people across directorates in Civic, Barton and Fyshwick, the arithmetic adds up fast.

What the Audits Are Actually Finding

Internal technical reviews conducted across three ACT directorates between January and May 2026 found that digital asset libraries had grown by an average of 40 percent over the preceding four years, but usable, unique image assets had grown by only about 22 percent over the same period. The gap — nearly 18 percentage points — is largely attributable to duplicate uploads, re-saves and format conversions that were stored rather than replaced. One directorate's content management system was found to contain more than 4,700 image files, of which a deduplication scan identified close to 1,600 as functionally identical to another file already in the system.

The University of Canberra's library and digital services team, which manages image assets for public-facing research and student platforms on the Bruce campus, ran a comparable internal audit in late 2025. Staff identified that their media repository held approximately 2,200 duplicate image entries out of roughly 9,000 total files — just under 25 percent. The exercise took three staff members around six working days to complete manually, a workload that prompted the team to begin evaluating automated deduplication tools in the first quarter of 2026.

The Australian National University, whose digital collections span the Acton campus and extend to the Noel Butlin Archives Centre in Dickson, has faced similar pressures. Long-running digitisation programs — particularly those tied to historical photograph collections — have generated large volumes of near-identical scans at different resolutions, all retained on the assumption that higher-resolution versions might eventually be needed. That assumption has a storage cost: high-resolution TIFF files can run to 80 megabytes per image, meaning even a few hundred duplicates consume tens of gigabytes.

The Cost Calculation — and What Comes Next

Cloud storage pricing for government and educational institutions in Australia typically falls between $0.02 and $0.05 per gigabyte per month under enterprise agreements, meaning a library carrying 500 gigabytes of duplicate data is paying somewhere between $120 and $300 per year just to store files it doesn't need. That sounds modest in isolation. Across a portfolio of eight to ten directorates, each carrying their own content management infrastructure, the cumulative waste becomes a genuine line item.

The more consequential cost is operational. Web platforms that pull images from bloated, poorly indexed repositories load more slowly, return more search noise and require more manual curation to maintain. For public-facing ACT government sites — including those serving residents in Gungahlin and Belconnen, where population growth has pushed up demand for online service delivery — that friction translates directly to user experience problems.

The ACT government's Digital Strategy, which runs to 2025 and has an anticipated successor framework, identifies data quality as a foundational priority. Duplicate image management sits within that remit, and several directorates are understood to be evaluating purpose-built deduplication software as part of their next procurement cycle, with decisions expected before the end of calendar 2026. For agencies still managing the problem manually, the ANU and UC audits offer a useful benchmark: set aside dedicated staff time, run a hash-based scan of the full asset library first, and budget for at least one full week of remediation work per 10,000 files held.