NAFULA: Kenya's health insurance sector is growing, just not profitable yet

Latest IRA report indicates that medical insurance grew by 22.4 per cent last year, becoming the largest class in general insurance, at 41.1 per cent of the segment.

Premiums grew 13.4 per cent in the first half of the year. New members are being onboarded. New products are being launched. By almost every headline measure, things are moving in the right direction.

And yet the combined ratio for general insurance sat at 108.3 per cent in that same period. Which means insurers collected premiums but then paid out more than they collected. That shifts the perspective as it’s not a growth challenge only but a structural problem as well.

We have watched this play out up close. As a technology platform sitting at the intersection of insurers, providers and members, we see what the headline numbers do not show: the claims that arrive weeks after treatment despite ability for real-time claims submission, healthcare providers not embracing these changes that come with digitisation from paper claim to e-claim.

The reconciliation that drags on. The fraud that gets caught, before and after the payment has left the door. The pricing decisions made on data that is months old by the time anyone looks at it, despite the ability to monitor real time.

The industry's response to all of this has been, for a long time, to keep growing. Bigger portfolio, new segments, more members.

The idea being scale would eventually fix the margins. It has not. In most cases, scale has just meant more of the same problem, at higher volume. Insurance is a risk management business. That sounds obvious, but it has an implication that the industry has not always lived up to, you cannot manage risk you cannot see.

For years, visibility into what was actually happening across the healthcare journey, consultations, diagnostics, pharmacy, hospital admissions, provider billing, was limited at best. Up to 50 per cent of data became inaccurate by the time it was analysed, with reporting timelines stretching from weeks to months.

Claims came in retrospectively. Fraud detection happened after the fact. Pricing was built on last year's averages, not this quarter's patterns.

That is not risk management. That is administration with a lag. And in a market where medical inflation keeps climbing and fraud cases tripled in just one quarter of 2025, the lag is costly.

Something is shifting, and the numbers are starting to show it. The digitisation of the administration has created something that did not exist before: a real-time data trail across the healthcare journey. Hospitals, clinics and pharmacies are now on connected billing systems.

Mobile connectivity means healthcare transactions can move onto platforms that all parties can see, in real time, combined with modern underwriting core platform as opposed to legacy platforms that could not connect data in real-time. When that infrastructure is combined with AI, the results are not theoretical. They are already happening here.

One insurer reported that AI has allowed 80 per cent of claims to be processed faster, without compromising accuracy. Another market leader blocked Sh400 million worth of fraudulent claims in 2025 through AI-enabled detection. These are not pilot results.

These are numbers from the market we are all operating in. And this is still early. A 2024 Deloitte survey found that only one per cent of insurance operations in Africa currently use AI, but that figure is projected to reach 80 per cent within five years.

The KPMG 2025 Africa CEO Outlook found that 41 per cent of African CEOs now rank AI integration as their top investment priority. The direction of travel is not ambiguous.

At M-Tiba, the shift is already operational. Our AI Decision Engine for pre-auths and claims now processes all requests in real time, without manual intervention, adjudicating against benefit rules at the point of service rather than days after.

Where clinical judgement or claims complexity demands it, a human assessor remains in the loop — the engine routes and flags, the expert decides. Our fraud detection models identify and flag suspicious claims five times faster than manual review, before payment leaves the system.

The combined result: 100 per cent of successful submission pass though this AI Decision Engine, shortening the time spent on a claim. Insurers using our platform have also seen healthcare costs fall by up to 15 per cent, with portfolio margin improvements of 10 to 20 percentage points.

These are not projections. They are what is already running on the Kenyan market.

What this actually changes is the business model. When claims can be adjudicated in real time against benefit rules, leakage drops. When provider billing patterns are monitored live, fraud gets caught before payment rather than after.

If after, recourse action is taken faster for recovery. When pricing is built on current utilisation data rather than historical averages, it reflects reality. When members can track their claims and get instant pre-authorisation responses, trust builds.

None of that is about technology for its own sake. It is about finally having the operational control that health insurance has always needed but rarely had. The insurers who are moving in this direction are not just improving their ratios.

They are changing what it means to run a health portfolio, from a line of business that is structurally difficult to one that can be actively and profitably managed.

As Kenyans we have done this before. Mobile money was not imported. It was built here. And it did not just change how Kenyans moved money, it rewired financial services across the continent and became a global reference point for what is possible when infrastructure, innovation and market need align.

The conditions for a similar shift in health insurance are present. The provider digitisation is underway. The modern platform infrastructure is built. The data is becoming available. And with insurance penetration still sitting at just 2.4 per cent of GDP, the market need is not going anywhere.

What has been missing is the operational layer that makes growth sustainable, not just possible.

That layer is now here. The next chapter of health insurance will not be written by whoever grows the fastest. It will be written by whoever executes with the most discipline. Whoever uses their quality data.

Whoever manages their portfolio actively rather than passively. Whoever stops treating profitability and coverage expansion as competing goals and starts pursuing them together.

That is what we are building toward. And based on what we are already seeing, it is closer than most people think.

Acting managing director at M-Tiba