Underwriting Is Racing Ahead. Total Loss Is Still Manual. Where That Leaves AMCs.

There is a quiet asymmetry opening up across UK motor insurance, and it is worth talking about. Underwriting is automating quickly. Agentic AI has moved from pilot to production. Cytora's deployment with Zurich has been publicly cited as proof that agentic AI is past the experimentation phase. Submissions, triage and decisioning at the front of the funnel are increasingly handled by tools that flow data straight through to a quote.

Total loss settlement has not moved at the same pace. The valuation conversation is still manual. The challenge cycle is still manual. The customer touchpoints are still handler-led. LexisNexis recently flagged the operational inefficiency this creates, and Insurance Post has been writing about straight-through processing as a strategic objective rather than a current reality.

The gap matters. For AMCs sitting between insurer underwriting and customer settlement, the operational pressure now comes from both directions. The submission you receive arrived through automation. The settlement you produce still leaves through a manual process. That is the asymmetry, and it is unlikely to resolve itself in the next twelve months.

Why total loss is still manual, honestly

Three reasons sit underneath the slow pace, and most of them are good reasons.

The first is asymmetric risk. A wrong underwriting decision costs you a sale. A wrong total loss settlement costs the customer their car replacement money, attracts an FCA complaint, or both. Consumer Duty applies sharply to total loss valuations after the September 2025 multi-firm review, and AMCs and insurers know it. The pace is cautious because the cost of getting it wrong is high.

The second is valuation complexity. A total loss valuation is not a single number that can be looked up in a table. It is the result of a comparator analysis, a PAV uplift assessment, a condition adjustment and sometimes a negotiation with the customer over disputed value. That work resists pure automation because it depends on file-specific judgement applied to a specific vehicle. Automation can support the work. It cannot do it end-to-end without a human in the loop.

The third is customer experience. A total loss is a stressful moment for a claimant. The vehicle they relied on is gone. The settlement they receive determines whether they can replace it. The human-handler touch matters here in a way it does not in a typical underwriting submission. Removing the handler entirely from the conversation is not the right answer.

Those are the reasons. None of them are excuses. They explain the pace without justifying complacency.

What the underwriting asymmetry actually does to AMC operations

The practical pressure shows up in four places.

Cycle time expectations creep up. When an insurer can issue a quote in seconds and an underwriter can clear a submission in minutes, the same insurer's claims team becomes less patient with a total loss file that takes four to six weeks to settle. The expectation drift is real even when the operational reality cannot match it.

Audit trail expectations rise. Automated underwriting captures every decision step in a clean log. Manual total loss work has historically captured a thinner trail. The Consumer Duty review and the FCA's continued focus on claims handling mean the audit standard for total loss is now what the audit standard for automated underwriting already was. The bar moved upward.

Customer comparison gets sharper. The same customer who bought their policy via ChatGPT in a thirty-second conversation now has a total loss claim that takes a month. The contrast between buying experience and claims experience widens, and the brand pays for the gap whether or not the operational reality justifies it.

Handler workload concentrates on the wrong parts. When the easy parts of a total loss file are no longer easy to do faster, the handler ends up doing the routine parts at the same speed as the contentious parts. That is the operational inefficiency LexisNexis identified.

Where automation does belong in total loss

This is not an argument against automation in total loss. It is an argument against pretending the valuation decision itself is automatable. Plenty of the surrounding work is.

Evidence capture is automatable. Comparator pulling is automatable. Document classification is automatable. Customer status updates are automatable. The PAV uplift table is automatable. The audit trail itself is automatable. None of these touch the valuation judgement. All of them remove hours from the handler's week and put those hours back where they belong, on the parts that need the judgement.

The right pattern for total loss in 2026 is not end-to-end straight-through processing. It is augmented manual settlement. The handler stays in the chair for the valuation conversation. The system handles everything around them.

That is the operational shape that matches the asymmetric risk reality, the regulatory environment, and the customer experience requirement at once.

What AMC total loss leaders should be doing now

Three practical actions, regardless of platform.

First, separate the parts of your total loss workflow that need human judgement from the parts that do not. Write them as two lists. The first list (valuation calls, customer-facing conversations, negotiation with insurers on disputed cases) stays human. The second list (comparator pulling, PAV documentation, audit-trail capture, customer status updates) is your automation target.

Second, measure the cycle time on each part separately. Most AMCs measure total loss cycle time as a single end-to-end number. The useful measurement is the breakdown. How long does each handler spend on judgement work versus admin work this week? The judgement work is what should be growing. The admin work is what should be shrinking.

Third, watch the audit trail standard. The post-September 2025 FCA expectation is that every total loss decision has a clean, defensible trail behind it. The manual standard for that audit trail is rising. Automating the trail itself is the lowest-risk and highest-return automation move available in total loss today.

Where TotalSettle fits

TotalSettle was built around the augmented manual settlement pattern. The handler stays in the chair for the valuation. The system handles the comparator pulling, the PAV documentation, the audit trail and the customer status updates. Every file leaves a clean, defensible record without the handler reconstructing it after the fact.

That pattern is the right operational answer for total loss in 2026. Underwriting will continue to automate faster. Customer expectations will continue to drift toward conversational speed. Total loss will continue to require handler judgement at the valuation step. The thing that decides whether the AMC closes the gap or widens it is what happens in the middle layer, where the routine work either supports the handler or buries them.

TotalSettle is the total loss settlement portal built for AMCs. Faster settlements, cleaner audit trails, fewer FCA headaches. Book a demo.