Settle in seven days, the regulator says. Get it right, the policyholder says.
Claims is where the promise is kept or broken, and where the insurer is now squeezed from two sides: a regulator counting the days to a decision, and a duty to decline only on grounds that will survive a dispute. Speed and fairness pull against each other, and the claim that is rushed or wrongly declined is the one that reaches the regulator.
Claims handling is the moment the policyholder finds out what the policy was worth, and the regulators have moved decisively to govern it. Kenya's Insurance Regulatory Authority has published draft Claims Management Guidelines requiring an insurer to acknowledge a claim notification within two working days and to communicate its decision within seven days of receiving the investigation report, explicitly to curb the discretion insurers have used to delay and reject. The trigger is visible in the data: declined claims rose to 1.51 billion shillings in the first half of 2025, up sharply on the year before. The regulator is now counting.
The conduct regimes elsewhere point the same way. South Africa's Treating Customers Fairly framework names claims handling as one of the six outcomes an insurer must demonstrate, and its courts have made clear that a cynical or thinly grounded repudiation is a bad-faith act — in disputes over business-interruption and misrepresentation claims, the conduct authority has acted against insurers declining where no proper ground existed. Nigeria's reform act states a zero-tolerance position on settlement delays. Across the continent, the leisurely or defensive decline is being supervised out of existence.
The squeeze this creates is real and structural. Speed and fairness are not naturally aligned. A seven-day clock pushes the claims function toward fast decisions, and the fastest decision is often the defensive decline or the under-settlement that protects the loss ratio. But a decline made to beat the clock, on grounds that do not hold, is precisely the decision that becomes a dispute, a complaint to the regulator, and — in the conduct framework — evidence of unfair treatment. The insurer that optimises only for speed manufactures the very conduct exposure the speed rules were meant to address.
Automation is the obvious answer to the clock, and the dangerous one if it is ungoverned. A model that auto-adjudicates claims can clear the queue inside the regulator's window, but a model that declines a claim and logs the action without the reasoning leaves the insurer unable to show, when the policyholder disputes it, why the decline was justified. The speed is real; the defensibility is absent; and in a conduct regime the undefended decline is worse than a slow one, because it is the pattern a regulator builds a case on.
The settlement clock also interacts awkwardly with the investigation it depends on. The rule runs from receipt of the investigation report, which creates an incentive either to rush the investigation or to park the claim indefinitely as one still under investigation — both of which the regulator is alert to. A claim referred to the fraud desk sits precisely at the seam between the settlement clock and the fraud process, and a referral made to stop the clock rather than because the evidence warrants it is its own conduct exposure. The honest claim parked as suspected fraud, to buy time against the deadline, is one of the most damaging things a claims function can do.
The asymmetry of the two errors is what makes this hard. Pay a claim that should have been declined and the insurer absorbs a loss. Decline a claim that should have been paid, under time pressure, on a ground that does not hold, and the insurer has a bad-faith repudiation, a complaint, and a conduct exposure that compounds across every similar decision. The two failures are not symmetric, and a claims process tuned only to turnaround time is tuned to the more dangerous one.
A decline made to beat the clock, on a ground that does not hold, is the cheapest decision today and the most expensive one when it reaches the regulator.
Where each sits.
Akki holds the claim record as a governed substrate — the policy, the notification, the investigation report, the correspondence — so a decision is made against one coherent record rather than fragments across systems, and reconstructed exactly if it is later disputed. When the regulator's clock is running, the information the decision needs is in one place, which is what makes a fast and defensible decision possible at once.
Solva structures the claims decision and refuses to ratify a decline the evidence does not support, surfacing what is missing rather than producing a fast repudiation that will not hold. Underneath each decision sits the basis — the policy terms applied, the facts relied on, the reasoning — which is the artefact a dispute or a conduct review reaches for. The speed comes from clearing the well-evidenced claims confidently; the protection comes from refusing to rush the ones that are not.
In most general claims handling SyniSense does little, and that is worth stating plainly: the data is the insurer's own and stays inside. Where claims involve sensitive health data — medical and disability claims especially — it keeps that identifiable detail inside the perimeter while the decision is reasoned. For routine motor or property claims, the weight sits with Akki and Solva.
For the claims manager, speed and fairness stop being a trade-off. The well-evidenced claims are decided confidently inside the regulator's window, and the ones that are not are held with the missing evidence surfaced, so the clock is met without manufacturing declines that will not survive a dispute.
For the policyholder, a decline comes with a reasoned basis grounded in the policy and the facts, not a defensive repudiation issued to beat a deadline. The fair-treatment outcome the conduct regimes demand is produced by the process rather than asserted in a policy document.
For the medical and disability claimant, whose claim turns on sensitive health evidence and often on urgent need, a grounded and timely decision matters most of all. The same coherent record that satisfies the regulator's clock lets the insurer decide a health claim on the evidence rather than delay it into a complaint — in precisely the class of claim where delay does the most human harm.
For the conduct regulator, the insurer can show that its claims decisions are timely and grounded — both halves of what the settlement rules and the fair-treatment outcomes require. A pattern of reasoned declines is a different supervisory picture from a pattern of fast ones nobody can explain.
For the board, the claims book stops accumulating undefended declines that a regulator could later assemble into a bad-faith pattern. The loss ratio is protected by getting decisions right, not by declining fast and hoping the disputes do not come. And the declines that are made are ones the insurer would be content to see read back to it by a tribunal — which is, in the end, the only test of a claims decision that matters.