London's Duplicate Image Problem: The Key Decisions That Will Shape the Capital's Visual Record

London's public institutions are sitting on a problem that has been accumulating for years. Duplicate and mislabelled images — photographs, planning maps, heritage records — have piled up inside digital archives managed by bodies including the London Metropolitan Archives on Northington Street and the Historic England offices in Waterloo. The question now is not whether to fix it, but who pays, who decides, and whether the deadline is realistic.

The issue landed back on agendas this summer partly because of the Starmer government's push to digitise planning records across English councils by the first quarter of 2027. That deadline, embedded in the Planning and Infrastructure Bill currently moving through Parliament, puts direct pressure on every London borough to verify the integrity of their digital document stores — and duplicate imagery is one of the most persistent and labour-intensive failures inside those stores.

Why the Backlog Is Bigger Than Anyone Admitted

A single planning application in a dense urban borough like Tower Hamlets or Southwark can generate dozens of site photographs, elevation drawings and aerial images. When those files are uploaded, renamed, or migrated between systems — as happened across most councils during the shift to cloud storage between 2019 and 2022 — duplicates multiply. The Greater London Authority's Digital Planning programme, which has been running since 2021, identified duplicate image records as one of the top three data quality failures in a review of borough submissions, though the precise count of affected files has not been made public.

The cost of human review is not trivial. Specialist archival data work in London currently runs at roughly £35,000 to £50,000 per year for a single qualified records officer, according to salary benchmarks published by the Archives and Records Association in its 2025 workforce survey. Smaller boroughs — Kensington and Chelsea, Richmond upon Thames — have fewer in-house staff to absorb that work. Larger ones, including Lambeth and Hackney, have more files to process in the first place.

Automated deduplication tools exist, and several councils piloted them through the London Office of Technology and Innovation's 2024 procurement framework. But automation alone does not solve the problem. Images that are visually identical but attached to different planning references, or photographs taken seconds apart that represent genuinely distinct records, require a human decision. Getting that wrong in a heritage or planning context has legal consequences, particularly where listed buildings on streets like Fournier Street in Spitalfields or Cheyne Walk in Chelsea are concerned.

The Decisions That Cannot Be Deferred

Three choices are now pressing. First, boroughs need to decide whether deduplication is treated as a one-off clearance exercise or embedded into ongoing upload protocols — the difference between a remediation project and a systemic fix. Second, the GLA and Historic England need to agree on a shared metadata standard so that an image tagged correctly in one system does not arrive as a duplicate when transferred to another. That negotiation has been ongoing since at least late 2024 with no publicly announced resolution.

Third, and most politically charged, is the question of central funding. The Department for Housing, Communities and Local Government has allocated resources to the broader digital planning reforms, but borough leaders have argued those funds do not cover the archival remediation work that the new standards implicitly require. Without ring-fenced money, the deduplication task is likely to be deprioritised in favour of frontline planning case work, particularly as councils face continued pressure on discretionary budgets.

The practical upshot for Londoners is this: planning applications, heritage consultations and freedom of information requests that rely on photographic records will continue to produce unreliable results until the underlying data is cleaned. For anyone pursuing a listed building consent near the Thames or challenging a development decision in Bermondsey, that is not an abstract concern. The window to get the governance right, before the 2027 digitisation deadline arrives, is narrowing. The conversations happening inside County Hall and the GLA's City Hall offices on the South Bank this autumn will determine whether London's institutions treat this as the infrastructure problem it is, or let it slide into the next cycle of emergency remediation.