The Hidden Cost of Duplicate Images: What the Numbers Tell Us About Canberra's Digital Storage Crisis

Canberra's public sector is drowning in duplicate images. Across Australian Public Service agencies headquartered in the capital — from the Department of Finance on King Edward Terrace to Services Australia's sprawling Tuggeranong complex — IT asset managers are grappling with digital storage portfolios bloated by years of unchecked file duplication, and the cost per gigabyte is no longer trivial enough to ignore.

The timing matters. The federal government's Digital Transformation Agency has been pushing agencies toward consolidated cloud storage frameworks since at least 2023, and the renewed pressure on departmental IT budgets in the 2025–26 federal budget cycle has forced a harder look at what exactly is consuming that storage. Duplicate image files — product photos, scanned documents, communications assets, satellite and geospatial imagery — routinely account for a disproportionate share of storage overhead in large organisations, according to IT asset management literature published by bodies including the Australian Signals Directorate.

At the Australian National University in Acton, the library and research computing teams have been working through a multi-year digital collections audit. ANU holds significant photographic and archival image collections, and deduplication work — systematically identifying and removing or consolidating identical or near-identical image files — has become a standing line item in operational planning. The University of Canberra at Bruce faces a comparable challenge across its creative and design faculties, where student project files and marketing asset libraries can multiply rapidly across shared network drives.

What the Storage Numbers Actually Look Like

Enterprise storage is not cheap. Commercial cloud storage pricing for large government contracts typically runs in the range of cents per gigabyte per month, but those cents compound fast. A mid-sized federal agency managing a 500-terabyte image archive — not an unusual figure for a department with geospatial, communications and records functions — could be carrying tens of terabytes of pure duplicates. Independent IT audits of comparable public sector environments in comparable jurisdictions have found duplication rates ranging from 20 to 40 percent of total unstructured data. Apply that to 500 terabytes and the redundant payload alone could exceed 150 terabytes.

Duplicate image replacement — the process of identifying redundant files, selecting a single canonical version, updating all references to point to that master, and deleting the rest — is not simply a storage efficiency exercise. The metadata consequences are significant. An image used across 14 separate web pages or internal systems, each carrying a slightly different filename or compression artefact, creates version-control headaches that compound during website migrations, Freedom of Information requests, and accessibility audits under the ACT's own Digital Accessibility Policy, which applies to ACT government websites and online services.

The ACT government's own ICT infrastructure, managed partly through Shared Services ICT based in Canberra's city centre, covers dozens of agency websites and intranet environments. A 2024 whole-of-government web consolidation push — part of the broader ACT Digital Strategy — specifically flagged content rationalisation as a target area, which in practice includes image asset libraries sitting on Content Management Systems used by agencies from Access Canberra to Transport Canberra.

The Clean-Up and What Comes Next

Deduplication tooling has matured considerably. Perceptual hashing algorithms — software that generates a fingerprint of an image's visual content rather than its raw file data — can now identify near-duplicates even when two files differ in resolution, compression level or colour profile. This matters because a government communications team might hold a 4K original, a web-optimised JPEG and a thumbnail, all of the same photograph, stored separately across three systems with no linking metadata.

For Canberra households and small businesses, the practical equivalent plays out in iCloud and Google Photos libraries: the average Australian smartphone user generated roughly 2,000 photos in 2024 according to industry estimates, and consumer deduplication apps have found duplication rates of 15 to 30 percent in typical libraries. The enterprise problem is simply that version multiplied by thousands of staff accounts and years of accumulated shared drives.

Agencies and institutions tackling this now are advised to begin with a storage audit that separates structured from unstructured data, apply automated perceptual hashing across image libraries, and establish a single digital asset management system — ideally one that enforces canonical file references at the point of upload rather than attempting a retrospective clean-up every few years. The ACT government's Digital Ready for Business program offers some guidance on digital asset governance for smaller organisations in the territory, though large-scale enterprise deduplication remains a specialist procurement exercise. Getting the numbers right upfront is, in this case, the entire point.