Paris Archives and Digital Teams Race to Fix a Flood of Duplicate Images in City's Heritage Database

A data cleanup effort spanning multiple city departments came to a head this week after administrators at the Direction des Affaires Culturelles de Paris identified hundreds of duplicate images clogging the municipal heritage image database — a problem that has slowed planning decisions in arrondissements from the Marais to the 13th, and created confusion on public-facing tourism portals tied to Paris 2024 Olympic legacy projects.

The issue matters now because the city is mid-way through a €4.2 billion Seine urban regeneration programme that relies heavily on digitised heritage records. Architects submitting renovation dossiers along the Quai de la Rapée and around the Bibliothèque nationale de France site at Tolbiac have reported delays since May when duplicate file entries began generating conflicting metadata, according to documents circulated internally and reviewed by The Daily Paris. With the Grand Paris Express's Line 15 South nearing its extended operational rollout, pressure on accurate urban documentation has rarely been higher.

How the Duplicates Accumulated

The problem traces back to a 2023 migration of legacy photographic archives into the city's Paris Musées open-access platform, which manages images from thirteen municipal museums including the Musée Carnavalet on Rue des Francs-Bourgeois and the Petit Palais on Avenue Winston Churchill. During the migration, a batch processing script failed to flag images already ingested under variant filenames, meaning thousands of photographs — particularly pre-Haussmann streetscape images and post-war reconstruction shots — entered the system twice or three times under different catalogue identifiers. Librarians at the Médiathèque Marguerite Duras in the 20th arrondissement flagged the anomaly to their supervisors in late May. By June, the problem had propagated into the regional planning database shared with the Île-de-France prefecture.

Duplicate image replacement — the technical process of identifying redundant files, selecting a canonical version, and systematically retiring the others while preserving all associated metadata — sounds straightforward. In practice, with a collection now exceeding 150,000 digitised items on the Paris Musées platform alone, it requires semi-automated tools cross-checked by human cataloguers. The Direction des Systèmes et Technologies de l'Information, the city's IT directorate based in the 17th arrondissement, deployed a deduplication script on Monday. As of Friday morning, the script had processed roughly 60 percent of the affected image sets.

Knock-On Effects Across the City

The consequences have not been purely administrative. Several tourism operators using the Paris Convention and Visitors Bureau's licensed image feeds discovered this week that promotional materials for Seine-Saint-Denis venues — drawn into the same digital ecosystem ahead of the Olympics legacy campaign — were displaying outdated or mismatched photographs. One heritage trail app covering the Canal Saint-Martin briefly showed a 1960s photograph in place of a contemporary image of the Hôtel du Nord on Quai de Jemmapes, a discrepancy noticed by users and reported on social media on Tuesday.

The city has not announced formal financial penalties or public accountability measures tied to the error, but internal guidance issued Wednesday set a July 18 deadline for the affected databases to be fully reconciled. Any planning dossier submitted to the Commission du Vieux Paris that references images logged between March and June 2023 is being flagged for a secondary verification check before sign-off.

For Parisians and professionals working with these systems, the practical advice this week is clear: if you are submitting a heritage renovation application or licensing images through the Paris Musées portal, confirm with the catalogue team on Rue Beaubourg that your image identifiers post-date the July 2026 cleanup batch. The directorate has set up a dedicated verification contact within its digital collections unit. Architects working the Grand Paris Express corridor projects have been told to expect a two-to-three week administrative delay on image-dependent submissions, with normal processing expected to resume by the end of July.