Sydney's Battle Against Duplicate Images Online: How It Stacks Up Against London, Seoul and São Paulo

Sydney's public archives, news organisations and cultural institutions are facing an accelerating problem with duplicate and misappropriated digital images — and a patchwork of responses from the State Library of NSW on Macquarie Street to the ABC's Ultimo headquarters is producing uneven results compared to coordinated national programs running in cities like Seoul and Amsterdam.

The issue has sharpened in 2026 as generative AI tools capable of producing near-identical image variants have proliferated, making it significantly harder for digital asset managers to detect when original photographs, artworks or archival scans have been copied, altered slightly and recirculated — sometimes with false attribution. For Sydney's institutions, which hold some of the most photographed urban and Indigenous heritage material in the Southern Hemisphere, the stakes are not abstract.

The timing matters. The NSW government's ongoing digitisation push under its cultural infrastructure strategy, which has accelerated since 2023, has moved hundreds of thousands of images onto public-facing servers. More assets online means more exposure. And with Sydney hosting the 2025–2026 cycle of the Asia Pacific Screen Awards and serving as a regional media hub, the city's content ecosystem draws more external traffic — and more bad-faith reuse — than at any point in the past decade.

What Sydney Is Actually Doing

The State Library of NSW has been running its Digital Preservation Program, which includes metadata tagging and hash-based duplicate detection for its Flickr Commons collections, since rolling out updated protocols in late 2024. The library's catalogue on Macquarie Street holds more than a million digitised items, and staff there have been using open-source perceptual hashing tools to flag near-duplicate uploads across partner institutions. The program is real, but it is not centralised.

In Pyrmont, the Australian Associated Press photo desk has implemented its own internal reverse-image scanning workflow, cross-referencing outgoing and incoming wire images against major stock databases. Meanwhile, the Museum of Applied Arts and Sciences at Ultimo relies on its own rights management system, separate from what the State Library uses, separate again from what Create NSW oversees through its grant-funded digital projects. The lack of interoperability between these systems is a recurring frustration for digital archivists working across more than one institution.

Compare that to Seoul, where the National Information Society Agency has operated a centralised government image registry since 2022 that links public broadcasting, national museums and municipal archives under a single deduplication API. Amsterdam's Rijksmuseum, which made its entire collection of 700,000 high-resolution images available under open licence, embedded watermark and metadata verification into the download pipeline itself — meaning every reuse carries traceable provenance data from the moment of download. Sydney has no equivalent single-point system.

The Numbers and What They Suggest

A 2025 report from the International Federation of Library Associations and Institutions found that institutions operating under fragmented deduplication systems reported duplicate or misattributed image incidents at roughly three times the rate of those using centralised registries. Sydney's institutions, operating in that fragmented mode, have no shared incident count — which itself is part of the problem.

In practical terms, a photograph of, say, Circular Quay shot by a named Sydney photographer in the 1970s and held in a public archive can be scraped, slightly cropped, and uploaded to a stock image site with no attribution in under five minutes using freely available tools. Detecting that reuse, under Sydney's current architecture, requires someone to notice it manually or for the photographer or estate to file a complaint. That is a slow loop in a fast-moving environment.

London's British Library moved to an AI-assisted rights-flagging system across its digital newspaper archive, the British Newspaper Archive, in early 2025, reducing manual review time by an estimated 40 percent according to the library's own published figures. No comparable Sydney institution has published equivalent benchmarks.

What comes next will depend partly on whether the NSW government's planned Digital Assets Framework — flagged in the 2025–26 state budget papers as under development within the Department of Creative Industries, Tourism, Hospitality and Sport — moves from policy language to funded infrastructure. Institutions and independent archivists working in the sector say the framework's practical scope, particularly whether it will mandate interoperability between cultural agencies, will determine whether Sydney closes the gap with its global peers or continues managing the problem institution by institution, corridor by corridor.