Joburg's Digital Archives Are Full of Duplicate Images — Here's What Officials, Experts and Key Figures Are Saying About Fixing It

The City of Johannesburg holds tens of thousands of digitised photographs — infrastructure surveys, heritage documentation, planning records — and a significant portion of that archive contains duplicate images that clog storage systems, slow retrieval and, in some cases, have led to incorrect records being used in official processes. The question of how to clean up and manage those files is no longer just a technical headache. It has become a governance issue, and voices across the city's professional and civic landscape are starting to say so out loud.

The timing matters. Johannesburg is mid-way through a broader digital modernisation push, part of the ANC-DA coalition government in Gauteng's commitment to improving service delivery through better data infrastructure. The Joburg Metrorail reform programme, upgraded traffic-management systems along Louis Botha Avenue and new smart-city pilots in the Sandton financial district have all generated enormous volumes of photographic and visual data since 2023. Without proper deduplication protocols, that data becomes a liability rather than an asset.

What the Professionals Are Saying

At the South African Society of Archivists, members have been discussing the duplicate-image problem in municipal contexts for at least two years. The society's annual conference, held in Pretoria in March 2026, included a session specifically on automated deduplication tools for public-sector digital repositories. Practitioners from organisations including the Wits Historical Papers Research Archive on Yale Road in Braamfontein described how legacy scanning projects from the early 2010s produced duplicate files at rates sometimes exceeding 30 percent of total ingested material — largely because different departments scanned the same source documents without coordinating with a central registry.

Digital preservation specialists working with the Soweto Heritage Trust, which manages photographic records tied to Orlando, Meadowlands and Diepkloof going back to the 1950s, have noted a related problem: duplicate images stored under different filenames often carry conflicting metadata, meaning a photograph might be tagged with two different dates or locations. For a heritage organisation whose credibility rests on the accuracy of its records, that is not a minor inconvenience. It directly affects the integrity of exhibitions, educational programmes and community restitution projects.

Information technology firms operating out of the Sandton CBD have been pitching AI-assisted deduplication software to Joburg municipal departments since at least late 2024. The pitch is straightforward: hash-based matching tools can identify visually identical or near-identical images across large repositories and flag them for human review, reducing manual labour costs substantially. One proposal circulated among City of Johannesburg IT procurement officers in February 2026 cited projected storage cost savings of between 18 and 25 percent for large municipal archives, though those figures have not been independently verified by the city's auditors.

Pressure from the Coalition Government

The ANC-DA coalition in Gauteng has made e-governance efficiency a stated priority for the 2025-2027 term. That means procurement and records-management decisions are now scrutinised more closely by DA oversight structures than they were under single-party administration. Members of Mayoral Committee subcommittees dealing with corporate and shared services have raised duplicate data management as part of broader audits of city IT spending. The Auditor-General's office flagged data management inefficiencies in its 2024-25 municipal audit cycle, though the specific details of those findings for Johannesburg have not yet been published in full.

Community-level pressure is also building. In Alexandra, volunteers digitising historical photographs of the township for a local museum initiative launched on London Road in 2025 found that donated image collections frequently contained duplicates — sometimes hundreds of copies of the same photograph that had been shared repeatedly across family WhatsApp groups before being submitted. Without a deduplication step early in the ingest process, the archive risks growing unwieldy before it even opens to the public.

For anyone managing a Johannesburg-based digital archive right now — whether a city department, a heritage trust or a community organisation — the practical advice from practitioners is consistent: implement a deduplication review at the point of ingest, not retrospectively. Cleaning up an archive of 500,000 images after the fact is dramatically more expensive than filtering duplicates before they enter the system. Gauteng's Government Information Technology Office is expected to publish updated digital records management guidelines before the end of the third quarter of 2026, which should give both public bodies and civil society organisations a clearer framework to work from.