Nairobi's Duplicate Files Threaten Kenya's E-Government Push

Nairobi's push to digitise land records, business permits and social registries has produced an unexpected problem: duplicate images clogging government databases, inflating storage costs and slowing the very services they were meant to speed up. The Kenya National Land Commission, which oversees the bulk of the capital's parcel records held at the Ardhi House complex on Ngong Road, has been quietly working through a backlog of replicated scan files that accumulated during a 2023-2024 digitisation drive. How the responsible agencies handle that cleanup will shape the credibility of Kenya's broader e-government agenda.

The timing matters. President William Ruto's administration is under sustained fiscal pressure from the IMF austerity programme, and every shilling spent on redundant cloud storage is a shilling not going to the Nairobi Metro commuter rail expansion or informal settlement upgrading in areas like Mathare and Mukuru. The Gen Z tax revolt of 2024 left the government with almost no political room to defend wasteful public IT spending, making the duplicate-image problem more than a technical embarrassment.

How the Duplicates Accumulated

Government digitisation projects rarely fail at the scanner. They fail at the database layer. When operators at multiple county offices, including the City Hall annex on City Hall Way and satellite land registries in Embakasi and Kasarani, uploaded scanned title deeds and survey maps without a shared deduplication protocol, identical files were logged under different reference numbers. The result is a single physical document appearing two, three, sometimes more times in the national registry system, each copy consuming cloud storage that the government pays for in US dollars under contracts with commercial providers.

The Kenya ICT Authority, which sits under the Ministry of Information and Digital Economy, sets the technical standards for government data management. Its 2022 National Digital Master Plan called for all ministries to adopt a unified metadata tagging system by the end of 2025. That deadline passed. Without consistent metadata, file origin, scan date, operator ID, parcel reference, automated deduplication tools cannot reliably flag which copy is the authoritative one and which should be deleted. Deleting the wrong file in a land records system carries legal consequences that no county registrar wants to own.

Ardhi House is not the only pressure point. The Huduma Centre on Moi Avenue, which processes national ID renewals and business name searches, draws on overlapping registries. Duplicate entries there have been linked to processing delays reported by applicants waiting beyond the standard five-working-day turnaround. The Nairobi City County's own revenue portal, used for single business permit renewals, has similarly surfaced duplicate applicant profiles that require manual reconciliation before payment can be confirmed.

The Decisions That Cannot Be Deferred

Three choices are now unavoidable. First, the Kenya ICT Authority must decide whether to mandate a single deduplication standard across all agencies or allow each ministry to procure its own solution, a fragmented approach that would recreate the problem within two years. Second, Ardhi House must determine who has the legal authority to mark a duplicate file as void: a registrar, an ICT officer, or a panel. Without a clear answer, files will sit flagged but undeleted indefinitely. Third, the National Treasury must decide whether the ICT Authority gets a dedicated remediation budget line in the 2026-27 estimates, currently being finalised ahead of the June 2027 fiscal year, or whether agencies absorb cleanup costs from existing allocations already squeezed by the IMF programme's expenditure ceilings.

The six months between now and December 2026 are the window. Kenya's next World Bank Digital Economy for Africa grant review falls in early 2027, and auditors will look at data integrity metrics. If the duplicate backlog is still unresolved, it weakens the government's case for continued concessional financing for digital infrastructure. The Silicon Savannah reputation, built partly on the M-Pesa ecosystem and iHub's legacy in Nairobi's Westlands district, depends on a functioning public data layer underneath it. Startups building property-tech, credit-scoring or logistics tools on top of government APIs cannot scale on records they cannot trust. The technical fix exists. The political will to fund and enforce it is the actual variable.