Canberra Leads Australia on Duplicate Image Scrubbing, But Lags Behind Amsterdam and Seoul

The ACT Government's digital records team quietly completed a first-pass audit of duplicate imagery across its public-facing platforms in May 2026, covering more than 340,000 assets stored across data.act.gov.au and the ACT Open Data Portal. The exercise flagged roughly one-in-five images as either exact duplicates or near-matches — a rate that digital archivists say is consistent with repositories that have grown organically over a decade without enforced taxonomy rules.

The timing matters. Across the OECD, the push to clean up government image libraries has accelerated sharply since 2024, driven by two converging pressures: the cost of cloud storage at scale, and the risk that duplicate or mislabelled public imagery gets ingested into AI training sets and produces compounding errors downstream. For a city where the Australian Public Service employs roughly 100,000 workers and where agencies publish enormous volumes of planning, transport and environmental photography, the duplicates problem is less about aesthetics and more about data integrity.

What Canberra Is Actually Doing

The ACT's approach runs through two main channels. The first is the Digital Canberra Action Plan, which since its 2023 update has included image deduplication as a line-item deliverable for agencies under the Chief Digital Officer's office in Civic. The second is a procurement arrangement with the National Archives of Australia, whose Mitchell facility in Hume handles physical and digitised asset management for several ACT directorates under a shared-services agreement.

The Australian National University's 3A Institute, based on the Acton campus, has been a quiet partner in developing the hashing protocols used to identify near-duplicate images — work that began as part of a broader responsible-data research program. The University of Canberra's News and Media Research Centre in Bruce has separately been tracking how duplicate imagery spreads through public information systems, though that work is focused on journalistic rather than government archives.

Together, these efforts put Canberra ahead of Brisbane, Adelaide and Perth, none of which have published comparable audit frameworks for their digital image holdings as of mid-2026. Sydney and Melbourne have active programs but both are embedded in larger, slower state-government bureaucracies that make cross-agency coordination harder to achieve quickly.

How the Global Benchmarks Compare

Amsterdam's city archive — the Stadsarchief — completed a full deduplication of its 1.2-million-image public collection in late 2024, using open-source perceptual hashing tools and a six-person specialist team over 18 months. The project reduced the active image catalogue by 31 percent and cut annual storage costs by approximately €140,000, according to the Stadsarchief's 2024 annual report. Seoul's Smart City Data Hub, which went live in 2023, baked deduplication into the ingest pipeline from day one, meaning duplicates are rejected at upload rather than cleaned up retrospectively — a fundamentally different architectural choice that most legacy systems cannot easily replicate.

Wellington, the closest administrative comparator to Canberra given its size and federal-capital function, completed a whole-of-government image audit in 2025 under its Digital Public Service programme. Wellington's published results put its duplication rate at 23 percent before remediation — almost identical to the ACT's preliminary finding — but Wellington has since cleared roughly 80 percent of those flagged assets.

The ACT is, by contrast, still in the remediation planning phase. The Chief Digital Officer's office is expected to release a deduplication roadmap later in the 2026-27 financial year. Whether agencies are required to act on it, or merely encouraged to, is still being worked through Cabinet processes.

For public servants working out of Dickson, Woden or the Barton precinct who manage image-heavy publications — think the Transport Canberra photo library covering the light rail network, or the Environment, Planning and Sustainable Development Directorate's aerial survey holdings — the practical upshot is straightforward: audit your own holdings now, before a whole-of-government standard is imposed. The ACT Government's Digital Strategy team has published a self-assessment guide on the data.act.gov.au portal, and the National Archives Mitchell facility runs quarterly workshops for ACT agency staff. The next session is scheduled for August 2026.