How Canberra's government agencies ended up drowning in duplicate images — and what's being done about it

Somewhere inside the sprawling digital infrastructure that supports the ACT government's network of agencies, the same photograph exists hundreds of times over. Sometimes it's a stock image of a Gungahlin streetscape. Sometimes it's a branded headshot from a 2019 public consultation. Sometimes it's simply a scanned form that nobody deleted. The problem of duplicate images — redundant, untagged, storage-hogging copies of visual files multiplying across shared drives and content management systems — has quietly grown into one of the more unglamorous IT headaches facing Canberra's public sector.

The issue has come into sharper focus in 2026 as agencies face pressure to rationalise their digital infrastructure ahead of a broader ACT government data governance review scheduled to report later this year. The review, which covers records management practices across directorates including Health, Education, and Transport Canberra, is examining how agencies store, tag, and retire digital assets — and duplicate imagery sits near the top of the problem list.

How the archive got so cluttered

The roots of the problem stretch back to the early 2010s, when individual ACT government directorates each began building out their own websites, intranet portals, and digital communications channels largely in isolation. The shift to the Canberra.gov.au unified web platform — rolled out progressively from around 2016 — brought many of those sites under one roof, but it did not automatically reconcile the image libraries sitting behind them. Files migrated. Originals stayed put. New uploads duplicated what was already there.

The Australian National University's digital records research group, based on the Acton campus, has documented similar patterns in large institutional environments, noting that content migration projects routinely leave behind shadow copies when source directories are not decommissioned properly. The University of Canberra's library and information management school has flagged the same structural issue in curriculum materials aimed at public sector records officers.

At the same time, the ACT public service expanded its communications capacity through the 2010s, adding social media managers, web editors, and graphic designers across directorates. Each new role brought new upload habits and, often, no centralised mandate to check whether an image already existed in a shared repository. Photographic assets from major infrastructure announcements — the light rail Stage 1 opening on Northbourne Avenue in April 2019 being one well-documented example — ended up stored separately across Transport Canberra, Chief Minister's directorate, and the ACT government's media unit, sometimes in three or four distinct versions at different resolutions.

Storage costs and the compliance pressure building now

Cloud storage is cheap by the megabyte but expensive at scale, and the ACT government's digital footprint is large. The National Archives of Australia's digital preservation guidelines, updated in 2023, require that agencies applying for records disposal authority demonstrate they can identify unique versus duplicate records — a requirement that turns what seemed like a housekeeping issue into a formal compliance obligation.

Several Canberra-based digital asset management vendors, whose clients include ACT government contractors operating out of the Woden Town Centre precinct and Civic office blocks on London Circuit, have reported increased tender activity in the first half of 2026 for tools specifically able to run deduplication audits on image libraries. The tools use hash-matching — essentially generating a unique fingerprint for each file — to identify exact and near-duplicate copies, then flag them for human review before deletion.

The practical stakes are real. A single high-resolution image can run to 20 or 30 megabytes. Multiply that by thousands of duplicates across a directorate that has been operating for a decade and the cumulative storage bill is not trivial. More importantly, duplicate images embedded in published web pages create version-control problems: when branding updates or accessibility requirements demand changes to an image, editors who don't know which copy is the authoritative one risk updating the wrong file and leaving outdated versions live.

For agencies working toward compliance with the ACT's updated Digital Accessibility Policy, which sets requirements for image alt-text and metadata completeness, an uncontrolled duplicate library makes systematic tagging effectively impossible.

Directorates that want to get ahead of the governance review would do well to start with a full audit of their content management system's media library, map which directories are still actively written to, and nominate a single source of truth for image storage before migration or consolidation begins. The boring work, as it turns out, has become unavoidable.