Hong Kong's Digital Archives Face a Reckoning Over Duplicate Image Crisis: The Key Decisions Ahead

Hong Kong's public institutions are sitting on a problem they can no longer defer. Duplicate images — identical or near-identical digital photographs stored redundantly across servers, cloud platforms and legacy hard drives — have quietly consumed storage capacity and inflated IT costs across the government, cultural sector and media industry. The immediate question is no longer whether to act, but how, and who pays.

The issue has sharpened because several major digitisation deadlines are converging this year. The Hong Kong Public Records Office, which operates under the Government Records Service in North Point, is midway through a multi-year push to migrate paper and film archives into searchable digital format. At the same time, institutions including the Hong Kong Museum of Art in Tsim Sha Tsui and the Hong Kong Film Archive in Sai Wan Ho have expanded their digital collections substantially since 2021. Each expansion multiplied the duplication problem: the same photograph of a 1970s typhoon shelter, or a 1997 handover ceremony frame, may exist in dozens of slightly variant file sizes and compression settings across multiple repositories.

The Cost of Doing Nothing

Storage is not cheap, and enterprise-grade cloud or on-premise solutions carry real dollar costs that compound annually. Industry pricing benchmarks for enterprise cold storage in the Asia-Pacific region have ranged from HK$0.08 to HK$0.25 per gigabyte per month depending on the provider and redundancy tier, meaning a collection running into hundreds of terabytes generates six-figure annual bills before any active retrieval or processing costs are added. For government departments already facing budget scrutiny following the February 2026 financial results, that line item is attracting attention from the Financial Services and the Treasury Bureau.

The duplication issue also carries a less obvious risk: version confusion. When a photograph exists in multiple variants — different crops, different colour corrections, different metadata tags — archivists and journalists retrieving images may unknowingly use an altered or lower-quality version as the definitive record. For a city whose historical documentation has been subject to heightened political sensitivity since the National Security Law took effect in June 2020, the integrity of the visual archive carries weight beyond the merely administrative.

What the Next Six Months Will Decide

Three decisions will define the trajectory of the clean-up effort through the end of 2026. First, the Government Records Service is expected to finalise procurement specifications for a deduplication and digital asset management platform before the third quarter ends. The contract, which insiders in the IT sector have described to trade publications as likely to fall in the HK$20 million to HK$50 million range for an initial three-year term, will set the technical standard that smaller agencies and statutory bodies are likely to follow.

Second, the question of cross-institutional data sharing remains unresolved. The Hong Kong Film Archive and the university libraries — notably those at the University of Hong Kong on Pokfulam Road and the Chinese University of Hong Kong in Sha Tin — each maintain photographic collections with substantial overlap. A shared deduplication protocol would reduce redundancy across the sector, but it requires agreement on metadata standards, access rights and, critically, on which institution holds the authoritative master file when duplicates are found.

Third, smaller newsrooms and heritage organisations that cannot afford enterprise solutions are watching to see whether the Innovation, Technology and Industry Bureau will extend any kind of subsidised access to a central digital asset management service. No announcement has been made, but the bureau's Digital Economy Development Committee has listed cultural data infrastructure among its working priorities for the second half of 2026.

Organisations that want to position themselves well should begin an internal audit now — cataloguing collection sizes, identifying the oldest or lowest-quality duplicates and documenting which files carry rights restrictions. That groundwork will be necessary regardless of which procurement path or government framework emerges. Waiting for a definitive policy before starting the audit will simply move the problem into 2027.