Poor records could undermine Kenya’s AI push, archivists warn ahead of 10th KARMA conference

Get breaking business news on WhatsApp

Updated 31 October 2026

Sponsored

Kenyan government offices and companies rushing to adopt artificial intelligence could end up with unreliable answers unless they first fix how they keep their records, the Kenya Association of Records Managers and Archivists (KARMA) has warned. The association says incomplete, inaccurate and poorly secured files will feed straight into AI systems and weaken decisions, accountability and public services.

The warning comes days before KARMA’s 10th Annual Records Management Conference, which runs from 2 to 6 November 2026 at PrideInn Paradise Beach Resort and Spa in Mombasa. The official opening is on 3 November.

“Only as reliable as the information”

“Artificial intelligence is only as reliable as the information on which it operates,” said KARMA Secretary General Collins Mutimba. “If records are incomplete, inaccurate, poorly organized or insecure, AI systems may produce unreliable results.”

Mutimba wants institutions to put data quality, digitisation and information governance first, before they switch on tools that depend on that information. This year’s conference theme, AI-Ready Records for Smart Content-Based Governance, reflects the same worry.

The point matters for business as well as government. The association says government decisions, business deals and public services all rely on accurate, easy-to-find records, so weak filing has an economic cost, not just an administrative one.

Paper files and patchy systems still common

Kenya has made progress with e-government services and electronic records, Mutimba said, but it still trails countries with mature digital records and joined-up information systems. He listed four problems: fragmented information systems, too little funding, skills gaps and continued reliance on paper records in both public and private institutions.

The gap shows up elsewhere. A recent survey found that 73% of Kenyan firms lack mature tech skills despite their digital push, while the Central Bank has identified nearly 300 possible AI uses, from inflation forecasts to satellite nowcasting, all of which need clean data. Security is another worry, with the regulator reporting 2.7 billion cyber threat events in three months, and the state’s own Konza data centre is near full.

For ordinary Kenyans, the effect of digitised records is already visible in services such as the ArdhiSasa land search, where the quality of the underlying records decides whether a search can be trusted.

What KARMA is asking for

The association wants more investment in records management, stronger policy frameworks and better recognition of records and archives professionals. It is also seeking closer work with government on training, digitisation and the responsible use of AI.

KARMA was set up in 2010. Over the past decade it says it has drawn up a professional code of ethics, a competency framework and records digitisation guidelines to build skills among its members.

For workers, the message is that data and records skills are becoming part of the AI story. See our comparison of AI training platforms for Kenyans, and jobinkenya’s guide to getting paid to train AI in Kenya.

FAQ

What is KARMA?

KARMA is the Kenya Association of Records Managers and Archivists, a professional body set up in 2010 for records managers, archivists and information professionals.

When and where is the KARMA 2026 conference?

The 10th Annual Records Management Conference runs from 2 to 6 November 2026 at PrideInn Paradise Beach Resort and Spa in Mombasa, with the official opening on 3 November.

Why do records matter for AI?

AI systems work from the information they are given. KARMA says incomplete, inaccurate or insecure records can lead AI tools to produce unreliable results.

What problems does Kenya face in records management?

KARMA lists fragmented information systems, inadequate funding, skills gaps and continued reliance on paper records across public and private institutions.