Nairobi's Digital ID Drive Hits a Wall: The Duplicate Image Problem and the Key Decisions Ahead

Kenya's national digital identification system is carrying a quiet but serious flaw: duplicate images — the same photograph linked to multiple identity records — have been accumulating inside government databases, threatening the integrity of a system that millions of Kenyans depend on for everything from mobile money access to hospital registration. The problem is not hypothetical. Administrators at Huduma Centres in Westlands and along Ngong Road reported a spike in cases this year where applicants were rejected or delayed because biometric checks flagged their photographs against existing records belonging to different people.

The timing is rough for the Ruto administration. The government has staked considerable political capital on the Maisha Namba programme, its unified digital ID framework intended to replace the old Huduma Namba scheme and consolidate records across the National Registration Bureau, the Kenya Revenue Authority, and the National Hospital Insurance Fund. With an IMF austerity programme squeezing the fiscal space and the Gen Z tax revolt still fresh in public memory, any perception that public money has been spent building a broken database will land hard.

How the Problem Compounds

Duplicate images enter the system through several routes. Manual data-capture errors during enrolment drives — many conducted in high-throughput locations such as Gikomba Market and the Eastleigh bus terminus — mean that one person can be enrolled twice, or that a low-quality scan is overwritten by a new scan that the algorithm then misidentifies as a different individual. Legacy records migrated from the old Integrated Population Registration System, which date back to the early 2000s, were not always cleaned before they were folded into newer databases. The National Registration Bureau has previously acknowledged, in parliamentary committee proceedings, that the transition created record-matching gaps, though the precise scale of duplication has not been made public.

The financial stakes are tangible. Kenya's digital ID infrastructure has attracted more than Ksh 4 billion in government capital expenditure across successive budget cycles, according to figures tabled in National Assembly budget documents. The Maisha Namba rollout was formally announced in 2023, with enrolment targets that cut across all 47 counties. By mid-2025, the programme had processed records for tens of millions of Kenyans, making a systematic deduplication exercise both urgent and logistically enormous.

The Decisions That Will Define What Comes Next

Three choices now sit on the desk of the Interior Ministry's directorate responsible for civil registration. First, whether to commission a full database audit using automated facial-recognition deduplication — a technically sound approach but one that requires licensing proprietary software or building local capacity at the Kenya ICT Authority's data centre on Waiyaki Way. Second, whether to open a public correction window through Huduma Centres, allowing citizens to self-report mismatches, which would be cheaper but slower and dependent on public trust that the Gen Z protests have partially eroded. Third, whether to pause new enrolments in dense urban areas such as Mathare and Korogocho — where community-based enrolment drives produced the highest duplicate rates — until the existing records are cleaned.

The Silicon Savannah tech community has a stake in the outcome. Startups building Know-Your-Customer products on top of the national ID infrastructure — several of them operating out of iHub on Ngong Road — cannot reliably verify customers when the underlying records are inconsistent. A prolonged unresolved state effectively taxes innovation without any formal legislation doing so.

Civic groups operating in Kibera and Mukuru, which have pushed for informal settlement upgrading to include formal address and ID registration, warn that residents in those areas are disproportionately affected. People who enrolled during community outreach campaigns organised by NGOs, rather than at formal Huduma centres, face the greatest risk of having unstandardised photographs that the algorithm flags.

The government has until the next parliamentary session, expected in late September 2026, to present a remediation plan or face renewed scrutiny from the Public Accounts Committee. Whatever path Interior Ministry officials choose, the window to act before public frustration hardens into another political flashpoint is measured in weeks, not months.