Zurich's Digital Archives Are Full of the Same Photo Twice — Experts Want That Fixed

Zurich's municipal digital archive holds tens of thousands of images — aerial shots of the Limmat corridor, construction-phase photographs from the Europaallee development, portraits from city council sessions going back to the early 2000s. A significant share of them, according to specialists who have reviewed the holdings, appear more than once. Sometimes many more times than once.

The problem is not unique to Zurich, but the city's particular combination of sprawling cantonal databases, a university research sector anchored by ETH Zurich, and a post-pandemic push toward e-government has brought it into unusually sharp focus this summer. Several institutions are now publicly or internally pressing for a coordinated deduplication strategy — a process of identifying and removing redundant image files — before further migration to cloud infrastructure locks in the mess permanently.

Why It Matters Now

The urgency is partly technical and partly financial. Stadtarchiv Zürich, the official municipal records office located on Neumarkt in the Altstadt, has been expanding its digital holdings as part of Switzerland's broader push toward open-government data. The Federal Archives in Bern set a 2025–2030 digitisation roadmap that requires cantonal and municipal bodies to meet interoperability standards. Redundant files complicate those handoffs, increasing both storage costs and the risk that researchers pull different versions of the same image and treat them as distinct records.

Professionals working in digital asset management describe duplicate images as one of the most persistent and underestimated problems in institutional archiving. Estimates from comparable European municipal archives — including those in Hamburg and Vienna — have suggested that between 15 and 30 percent of stored image assets in large civic databases are functional duplicates, meaning identical or near-identical files that entered the system through separate upload channels over the years. Zurich has not published its own figure, but librarians and IT staff familiar with the Stadtarchiv's holdings say internal audits have flagged the issue repeatedly.

ETH Zurich's Chair of Information Science, based on Rämistrasse, has been developing machine-learning tools capable of detecting near-duplicate images even when file names, metadata and upload dates differ — the kind of sophisticated variation that trips up simple hash-based deduplication. Researchers there have described the civic archive sector as an important real-world testing ground for those tools, though no formal partnership with Stadtarchiv Zürich has been announced.

Pressure From Multiple Directions

The Zentralbibliothek Zürich on Zähringerplatz, which manages a parallel set of digitised historical image collections, has been working since 2023 on its own internal deduplication protocols. Librarians there have noted that images donated or transferred from third parties — estates, universities, corporate collections — frequently arrive with no deduplication pre-processing, meaning the same photograph can exist in several collections simultaneously with conflicting rights metadata attached.

That rights question adds a legal dimension that pure IT solutions cannot resolve on their own. Swiss copyright law, specifically the revised Urheberrechtsgesetz that came into force in April 2020, extended protections in ways that make it more complicated to simply delete a duplicate file without first verifying which instance carries the authoritative licence or provenance record.

City councillors on the Gemeinderat's culture and digitalisation committee have been briefed on the issue, though no specific budget line has been publicly allocated for a deduplication programme as of July 2026. A cantonal working group is understood to be reviewing options, including whether to adopt a centralised platform or give individual institutions discretion over their own workflows.

For now, the practical advice from digital archivists is straightforward: institutions uploading new image batches should run deduplication checks before ingest, not after. The cost of fixing the problem at the front end is substantially lower than retrofitting a solution across legacy holdings that stretch back 25 years. Several Swiss cultural institutions have already moved in that direction. Whether Stadtarchiv Zürich follows before the next major migration deadline — provisionally scheduled for 2027 — will determine how much remedial work lands on whoever manages the project next.