Zurich's Digital Archive Crisis: The Key Decisions Ahead as Duplicate Images Swamp City Records

Zurich's public institutions are sitting on a problem that has quietly compounded for years: vast libraries of duplicate digital images embedded in everything from the city's urban planning archives on Amtshaus IV to the photographic records held at the Stadtarchiv on Neumarkt. The duplication crisis has reached a point where administrators, archivists, and IT procurement officers can no longer defer a decision about what comes next.

The issue crystallised this spring when a cross-departmental working group convened by Stadt Zürich's Informatik-Dienste identified redundancy rates in certain shared document repositories that made routine retrieval slower and storage costs measurably higher. The group's internal report, circulated in May 2026, did not produce a resolution — it produced a list of open questions that now sit on the desks of department heads across Stadthaus.

Why This Moment Matters

The timing is not arbitrary. Zurich is mid-way through a digitisation push tied to the Smart City Zürich strategy, which targets fully interoperable city data systems by 2028. Duplicate image files are not merely an aesthetic annoyance; they inflate cloud storage invoices, slow AI-assisted search tools being piloted at the Zentralbibliothek on Zähringerplatz, and create legal ambiguity about which version of a planning document or architectural photograph is the authoritative one. In a city where direct democratic processes can trigger referenda over infrastructure decisions, having conflicting document versions in circulation is a governance liability, not just an IT headache.

ETH Zurich's Data Archive Services group, which manages research data for hundreds of active projects, has dealt with adjacent problems in academic repositories for over a decade. The university's RDMS guidelines recommend deduplication at ingestion, a standard that municipal systems have historically not enforced. The gap between academic best practice and city hall's actual workflow is now the central friction point.

Costs are real and rising. Cloud object storage priced at the tiers used by Swiss public-sector entities — typically between CHF 0.02 and CHF 0.04 per gigabyte per month under framework agreements negotiated through the federal procurement office — becomes significant when repositories run into tens of terabytes of redundant files. The Stadtarchiv alone manages holdings that have grown substantially since its digitisation programme accelerated after 2020.

The Decisions That Cannot Wait

Three choices now sit directly in front of city administrators. First, who owns the deduplication decision: the individual department that created the files, the central IT directorate, or an independent records officer with binding authority? The working group's May report flagged this governance gap as the single largest obstacle to progress. Without a clear mandate, every department defaults to keeping everything, on the grounds that deletion carries political risk.

Second, what technical standard gets adopted? Two competing approaches are under informal discussion. Hash-based deduplication — comparing file fingerprints to identify exact copies — is fast and relatively cheap to implement. Perceptual hashing, which catches near-duplicate images that have been resized or slightly recoloured, is more thorough but requires specialist software and longer processing time. The Stadtarchiv's professional archivists have reportedly argued for the more rigorous method, concerned that a hash-only pass would leave thousands of near-duplicates untouched.

Third, what happens to files flagged for deletion before they are removed? Swiss archival law under the Archivierungsgesetz establishes minimum retention periods for public records, and any automated purge process has to be certified as legally compliant. A misstep here could expose the city to challenge under cantonal administrative law.

A decision framework is expected to go before the relevant city committee before the summer recess ends in mid-August. If approved, a pilot deduplication run covering the Stadtentwicklung department's image library — one of the largest and most cited internally — could begin in September. Broader rollout across Zürich's shared drives would follow in 2027, assuming the pilot holds up under legal review. Residents and civic groups who interact with city records through the Öffentlichkeitsprinzip public access provisions will want to watch whether the process includes any external audit mechanism. Right now, it does not.