São Paulo Races to Purge Duplicate Images From City Digital Archives — and the Gap With Global Peers Is Closing Fast

São Paulo's municipal digital infrastructure team quietly hit a milestone last month: the Prefeitura's open-data portal, hosted through the SP Dados platform on Viaduto do Chá's administrative corridor, completed the first systematic audit of duplicate images across its urban planning and public works databases. The audit flagged tens of thousands of redundant files — aerial drone shots of Avenida Paulista, flood-damage photographs from the Tietê River basin, and construction-site records from Cohab projects in the East Zone — that had been stored in multiple copies across disconnected government servers for years.

The problem sounds mundane. It isn't. Duplicate image files inflate storage costs, slow emergency-response systems that rely on real-time visual data, and create legal liability when conflicting versions of the same urban-monitoring photograph are submitted as evidence in licensing disputes or environmental hearings. With São Paulo managing one of the densest urban surveillance and civic documentation systems in Latin America — covering 96 subprefeituras and more than 12 million residents — the administrative overhead of image redundancy translates directly into public money and institutional dysfunction.

What São Paulo Is Actually Doing

The city's approach is anchored in a partnership between the Secretaria Municipal de Inovação e Tecnologia and two local startups from the SP Tech ecosystem based in Vila Olímpia. The program, operating under the broader Smart Sampa digital governance initiative launched in early 2025, uses perceptual hashing algorithms — a technique that generates a compact fingerprint for each image, allowing near-identical duplicates to be flagged without manual review — to process incoming visual files before they enter the permanent archive. The technology is not novel globally, but São Paulo's implementation is notable for applying it retroactively to legacy databases rather than only at the point of new intake.

The Arquivo Histórico Municipal on Rua Roberto Simonsen in the city centre is running a parallel humanities-focused version of the same problem: thousands of digitised historical photographs from the 1950s and 1960s urban expansion era exist in multiple scans of varying quality. Archivists there have been working since March 2026 to reconcile duplicates by hand, a slower process that the tech secretariat's automated tools are not yet equipped to handle, given the sensitivity of heritage materials requiring curatorial judgment.

How São Paulo Compares to Other Cities

The comparison with other major cities is instructive. Mexico City's CDMX Digital agency launched a similar image-deduplication initiative in 2024 for its metropolitan transit authority's camera archive, reportedly cutting storage costs by roughly 30 percent within the first year, according to reporting by Mexico City–based outlet Expansión. Bogotá's Secretaría Distrital de Planeación began automated visual-database hygiene protocols in January 2025, focusing initially on cadastral imagery used for property tax assessments in Chapinero and Usaquén. Both cities, like São Paulo, face the compounding challenge that their most critical visual data — flood monitoring, infrastructure inspection, protest-crowd documentation — is generated fastest during crises, precisely when staff capacity to manage duplicates is lowest.

London's Government Digital Service published guidance in 2023 recommending that all UK local authorities adopt content-addressable storage for new image ingestion, effectively making duplication structurally impossible at the upload stage rather than catching it afterward. São Paulo is not yet at that architectural level, but the SP Dados team has reportedly identified content-addressable storage as the medium-term target for the next phase of the Smart Sampa infrastructure upgrade, expected to be tendered in the second half of 2026.

For residents and businesses that interact with São Paulo's public image repositories — urban developers pulling environmental-impact records, journalists requesting public-works documentation under the Lei de Acesso à Informação, or community groups in Paraisópolis mapping flood vulnerability — the practical upshot is already tangible. Response times on image-based information requests from the Prefeitura's ouvidoria have been described in internal briefing documents as a priority metric for 2026. Whether the audit findings translate into a formal published report accessible to the public will depend on whether the secretariat treats the data as an operational internal matter or as open-government disclosure. Advocates from Transparência Brasil have previously argued that infrastructure audits of this kind should be published in full on SP Dados. That argument is still unresolved.