---
title: "Insurance claims automation in 2026: what's automated, what's still manual"
description: "Claims automation has reached FNOL, triage, estimating, payments, and detection. Investigation of flagged claims is the last manual step. Here is the 2026 stack map."
date: "2026-06-09"
lastModified: "2026-06-09"
author: "Pankaj Dhariwal"
tags: ["Guides"]
canonical: "https://gethesperai.com/blog/insurance-claims-automation-2026-whats-automated/"
---

# Insurance claims automation in 2026: what's automated, what's still manual

> **TL;DR** By 2026 automation owns the front of the claims pipeline - intake, triage, straight-through processing, damage estimating, payments, and fraud detection all have named software vendors. The one stage with no software incumbent is the deep investigation of flagged claims, which stays a manual SIU workflow at most carriers. That is the slice autonomous investigation agents are built to close, lifting coverage from roughly 25% to 100% of flagged claims.
>
> - 78% of P&C insurers adopted generative AI, only 4% scaled it in claims
> - Investigation is the last layer with no software vendor except manual SIU
> - Coverage from ~25% to 100% is the next loss-cost lever, not more front-end automation

- **82%** - Carriers using AI somewhere in operations (Sedgwick report, via Risk & Insurance)
- **4%** - P&C insurers that scaled genAI in claims (Bain & Company survey, via Risk & Insurance)
- **~25%** - Of flagged claims investigated manually (The rest are paid, denied, or queued)
- **2-4 hrs** - AI investigation time per flagged case (vs 14+ days manual)

## Claims automation in 2026 is real and uneven

Claims automation has genuinely arrived, but it is concentrated at the front of the pipeline and thin everywhere judgment is required. For a claims-transformation owner, the practical question in 2026 is no longer whether to automate - it is which stage is left to automate, because the easy wins at intake, triage, and estimating are mostly booked. The harder, higher-value stage is the one that has lagged the rest: investigation of flagged claims.

The adoption data is broad and the scaling data is narrow, and the gap between them is the story. A Sedgwick report, [reported by Risk & Insurance](https://riskandinsurance.com/ai-adoption-in-property-claims-remains-fragmented-despite-rapid-growth/), found 82% of carriers now use AI tools somewhere in their operations, yet only 7% of property carriers have achieved scalable AI success. A Bain & Company survey of 160 global insurers, [also reported by Risk & Insurance](https://riskandinsurance.com/generative-ai-in-insurance-claims-faces-a-scale-problem/), found 78% of P&C insurers have adopted generative AI but only 4% have scaled it meaningfully across claims operations, and just 27% are pursuing comprehensive claims transformation.

Read those numbers together and the pattern is clear. Most carriers have AI running on discrete, high-volume tasks - data extraction, documentation, low-severity claims - rather than re-engineering the harder stages. The loss-ratio impact has landed where the work is repetitive. The stages that demand judgment, including the investigation of suspicious claims, are where adoption stalls. This post maps the full claims-automation stack layer by layer, places the named vendors at each layer, and isolates the one layer that still has no software incumbent.

If you want the full definition of the stage at the end of that map, our pillar on [autonomous AI claims investigation](/blog/autonomous-ai-claims-investigation-pillar) covers what it is and how it differs from detection.

## Map the pipeline: what's automated today

Walk the claims lifecycle stage by stage and the automation footprint is easy to trace. Each stage has matured at a different rate, but the front three-quarters of the pipeline now run with meaningful software automation at most mid-market and national carriers.

### FNOL and intake

First notice of loss and intake are largely automated. The Sedgwick report found intake automation cut average processing time from 10 days to 36 hours. Mobile-first FNOL and digital intake run on platforms like Snapsheet and the FNOL modules inside [Guidewire ClaimCenter](https://www.guidewire.com/products/claimcenter) and Duck Creek Claims, which capture the loss, validate the policy, and open the case file without a human keying it in.

### Triage and straight-through processing

Triage and straight-through processing now auto-approve low-complexity claims. Crawford & Company's Joel Raedeke, SVP of U.S. technology, told [Claims Journal](https://www.claimsjournal.com/news/national/2026/03/09/336163.htm) that "as AI drives more claims automation, we will see more straight-through processing of low complexity claims in 2026," with simpler claims passing through all decision gates to auto-approval. The Sedgwick data backs this up: low-severity claims now process up to 80% faster with AI in the loop.

### Damage estimating

Damage estimating is automated through computer vision and structured estimation utilities. Tractable's visual damage assessment compressed cycle time from days to minutes on specific claim types in its [Tokio Marine deployment](https://www.insurancejournal.com/news/international/2020/05/11/568090.htm), and Verisk's [Xactimate](https://www.verisk.com/solutions/claims/estimation-valuation/property-estimation-restoration/) remains the property-estimate standard. The Sedgwick report attributes up to 54% efficiency gains to AI photo analysis. An estimate, though, answers "how much" - not "is this claim legitimate."

### Payments and routing

Payments and routing are increasingly touchless. Once a low-complexity claim clears triage and estimating, the payout and case routing complete without manual handling. Across these four stages, the through-line is the same: automation has scaled where the task is high-volume and rules-friendly. None of these stages, by design, investigate whether a flagged claim is fraudulent.

## Detection is automated - and still not investigation

Fraud detection has its own mature automation layer, and it is easy to mistake a flag for a resolution. Detection scores or flags suspicious claims after FNOL, and the vendors here are well established. This layer scales. The question it leaves open is what happens to the flag.

Shift Technology's deployment at [AXA Switzerland](https://www.shift-technology.com/resources/case-studies/axa-switzerland-insurance-fraud-detection-success) screened over 1M claims and helped stop over EUR12M in fraud, which shows detection works at scale. FRISS and [Verisk ClaimSearch](https://www.verisk.com/products/claimsearch/) operate in the same layer with cross-carrier data and scoring. What none of them do is resolve the flag. They produce a signal that something is worth a closer look, and then a human SIU team takes it from there.

> **A flag is not a resolution**
>
> Rules-based detection carries a 60-85% false-positive rate. That means most flags are not fraud, and the only way to separate the real cases from the noise is to investigate each one. Detection is upstream; investigation is downstream. The detection layer tells you where to look. It does not do the looking.

With roughly 10% of property-casualty losses involving fraud and an estimated $308 billion lost annually in the US, per the [Coalition Against Insurance Fraud](https://insurancefraud.org/fraud-stats/), the volume of flags a mature detection layer produces is large. Every one of those flags is an investigation that someone has to run, or quietly drop. That is where the automated pipeline ends and the manual one begins.

## Investigation is the last manual step

The one stage automation has not reached at scale is the deep investigation of flagged claims, and the industry's own data admits it. The Sedgwick report is explicit that AI works most effectively on "high-volume, repetitive tasks - data extraction, documentation, low-severity claims handling and intake validation," and that 75% of claims professionals believe AI requires human oversight. Investigation is the opposite of high-volume and repetitive. It is low-volume, high-judgment, and case-specific, which is exactly why only 4% of carriers have scaled genAI in claims and only 7% have achieved scalable AI success.

For an SIU director, the consequence shows up as a coverage gap, not a productivity statistic. A manual SIU investigation takes 14+ days per case, and a single investigator carries 200+ open cases. The arithmetic does not close: with that caseload and that cycle time, manual SIU teams across US P&C carriers investigate roughly 25% of flagged claims. The other 75% are paid without full work, denied without full work, or queued indefinitely. The detection layer keeps producing flags; the investigation layer keeps absorbing only a quarter of them.

This is not a tooling failure inside SIU - it is a structural one. We covered the mechanics of it in detail in [why flagged insurance claims never get investigated](/blog/why-flagged-insurance-claims-never-investigated). The short version: the bottleneck is human attention, and no amount of front-end automation relieves it, because front-end automation produces more clean throughput, not more investigated flags.

> Automation moved front-to-back across the claims pipeline and stalled at investigation. The detection layer got faster at producing flags. The layer that turns a flag into a documented decision stayed exactly as constrained as it was a decade ago.
>
> - Hesper AI product research

## Why investigation resisted automation

Investigation stayed manual because it is not a single repetitive task that an automation script can wrap. FNOL, triage, and estimating each reduce to a well-defined operation a rules engine or a vision model can own. A fraud investigation is 15+ distinct judgment-heavy phases per case - document forensics, OSINT, statement cross-referencing, timeline reconstruction, financial pattern analysis, network checks - each producing evidence the next phase depends on.

Manual SIU teams run those phases one case at a time because a human investigator's attention is the bottleneck. You cannot run document forensics and OSINT and timeline reconstruction simultaneously when there is one person doing all of them. That serialization is the reason a case takes 14+ days and the reason coverage caps near 25%. The constraint was never that the work could not be defined - it was that one human could only do one phase at a time.

Autonomous investigation agents change the mechanism. They run all 15+ phases in parallel on every flagged claim, because an agent's per-case attention is not a shared resource. We break down how this works in [parallel processing of SIU investigation phases](/blog/parallel-processing-siu-investigation-phases). The result is order-of-magnitude compression: 14+ days drops to 2-4 hours, coverage moves from roughly 25% toward 100% of flagged claims, and per-case cost falls from about $2,500 to about $150. The investigator's role shifts from execution to decision-making - reviewing, overriding, and signing the audit-ready output rather than assembling it by hand.

## The 2026 claims-automation stack, layer by layer

Put every layer on one map and the gap becomes impossible to miss. Each layer has a job, a set of named automation vendors that occupy it in 2026, and a maturity level. Across FNOL, triage, estimating, payments, and detection, there is a named software vendor at every layer. At the investigation layer, the only incumbent is the manual SIU team.

| Layer | What it does | Automation vendors (2026) | Maturity |
| --- | --- | --- | --- |
| FNOL / intake | Capture loss, validate policy, open case | Snapsheet, Guidewire, Duck Creek | Mature |
| Triage / STP | Route and auto-approve low-complexity claims | Guidewire, Duck Creek, carrier ML | Maturing |
| Estimating | Assess damage and price the loss | Tractable, Verisk Xactimate, CCC | Mature |
| Payments | Issue payout, close routine cases | Carrier core + payment rails | Maturing |
| Detection | Flag and score suspicious claims | FRISS, Shift, Verisk ClaimSearch | Mature |
| Investigation | Resolve a flagged claim end-to-end with audit-ready report | Hesper AI (only software incumbent; else manual SIU) | Newly automatable |
| Adjudication | Final pay / deny / SIU decision | Human SIU lead and adjuster | Human-owned by design |

This is the point a detection vendor's byline cannot make. FRISS, Shift, and Verisk sit in the detection row, and their layer is mature - that is a fact in their favor, not a criticism. Tractable and Xactimate sit in estimating. Guidewire, Duck Creek, and Snapsheet own claims management and FNOL. Every one of those layers is occupied by software. The investigation layer - taking a flagged claim and resolving it end-to-end with a documented, audit-ready report - has no named software incumbent. Its incumbent is the SIU team at every carrier, working one case at a time. That is the layer Hesper is purpose-built for, and it is the only layer on the map where the answer to "which vendor automates this" was, until recently, "nobody."

> **Investigation is not adjudication**
>
> Hesper investigates the flagged claim and produces the audit-ready report. The human SIU lead and adjuster make the pay-or-deny call. The agent does not adjudicate, and it is not generic touchless automation - it sits downstream of detection and upstream of the human decision.

## What this means for a 2026 roadmap

If you have automated FNOL through payments and bought a detection platform, your next marginal loss-cost lever is the investigation layer, not more front-end automation. The front of the pipeline is where the easy efficiency lived, and the data shows it is largely booked: intake is down from 10 days to 36 hours, low-severity claims process 80% faster, photo analysis lifts handling efficiency by up to 54%. Squeezing those further yields diminishing returns. The un-automated stage is where the next basis points of loss ratio are sitting.

The coverage math is the lever. Only 27% of insurers are pursuing comprehensive claims transformation, and most of that 27% are still leaving investigation out of scope. With roughly 10% of P&C losses involving fraud and $308 billion lost annually, the gap between investigating 25% of flagged claims and investigating 100% of them is the single largest un-pulled lever in the fraud stack. No prior generation of fraud-tech moved it, because every prior generation operated upstream of investigation.

| Flagged-claim investigation coverage: manual vs autonomous | Value | Share |
| --- | --- | --- |
| Manual SIU (200+ cases each) | ~25% | 25% |
| Autonomous investigation agents | 100% | 100% |

The unit economics make full coverage affordable rather than aspirational. At about $2,500 per manual investigation, investigating every flagged claim was never financially possible - which is exactly why coverage settled at 25%. At about $150 per AI-investigated case, full coverage stops being a budget problem and becomes a workflow decision. When you are ready to evaluate the vendors at each layer side by side, our [AI fraud platforms compared](/blog/ai-fraud-platforms-compared-2026-pillar) breakdown maps detection and investigation tools against each other, and our [top fraud detection platforms guide](/blog/top-fraud-detection-platforms-2026) covers the detection layer in depth.

## Key takeaways

- By 2026, automation owns the front of the claims pipeline - FNOL, triage, straight-through processing, estimating, and payments all have named software vendors.
- Adoption is broad but shallow: 82% of carriers use AI somewhere, yet only 4% have scaled generative AI in claims and only 7% report scalable AI success.
- Fraud detection is its own mature layer (FRISS, Shift, Verisk), but a flag is not a resolution, and rules-based detection carries a 60-85% false-positive rate.
- Investigation is the last manual step: manual SIU teams reach only about 25% of flagged claims at 14+ days and ~$2,500 per case.
- Autonomous investigation agents run 15+ phases in parallel, cutting cycle time to 2-4 hours and lifting coverage toward 100% at about $150 per case.

## Frequently asked questions

### What parts of insurance claims are automated in 2026?

By 2026, automation covers the front of the claims pipeline. First notice of loss and intake are largely automated - Sedgwick reports intake automation cut average processing from 10 days to 36 hours. Triage and straight-through processing now auto-approve low-complexity claims; Crawford & Company expects more simple claims to pass through all decision gates to auto-approval. Damage estimating is automated through computer-vision tools, and AI photo analysis has lifted handling efficiency by up to 54%. Payments and routing are increasingly touchless. What is not automated at scale is the deep investigation of flagged or suspicious claims, which remains a manual SIU workflow at most carriers and is the gap autonomous investigation agents are built to close.

### How many insurers have actually scaled AI in claims?

Adoption is broad but scaling is narrow. A Bain & Company survey of 160 global insurers found 78% of P&C insurers have adopted generative AI, yet only 4% have scaled it meaningfully across claims operations, and just 27% are pursuing comprehensive claims transformation. A separate Sedgwick report found 82% of carriers use AI somewhere in their operations but only 7% of property carriers have achieved scalable AI success. The takeaway: most carriers run AI on discrete, high-volume tasks - data extraction, documentation, low-severity claims - rather than re-engineering the harder, judgment-heavy stages. Investigation of flagged claims is one of those harder stages, which is why it has lagged the rest of the pipeline.

### Why is claims investigation still manual when everything else is automated?

Because investigation is not a single repetitive task. FNOL, triage, and estimating are high-volume, rules-friendly steps, exactly where AI scales, per Sedgwick's finding that AI works most effectively on high-volume, repetitive tasks. A fraud investigation is the opposite: 15+ distinct judgment-heavy phases per case, including document forensics, OSINT, statement cross-referencing, timeline reconstruction, and network analysis. Manual SIU teams run those phases one case at a time because a human investigator's attention is the bottleneck, so they reach only about 25% of flagged claims and take 14+ days each. Autonomous investigation agents change the mechanism by running all 15+ phases in parallel on every flagged claim, compressing 14+ days to 2-4 hours.

### Is claims automation the same as fraud detection?

No - they are different layers. Fraud detection scores or flags suspicious claims after FNOL; vendors like FRISS, Shift Technology, and Verisk operate here, and that layer is mature. AXA Switzerland's Shift deployment screened over 1M claims and stopped over EUR12M in fraud, which shows detection scales well. But detection produces a flag, not a resolution, and rules-based detection carries a 60-85% false-positive rate, so flags still need investigation. Claims automation more broadly covers intake, triage, estimating, and payments. Investigation - taking a flagged claim and resolving it end-to-end with an audit-ready report - is its own downstream layer. Detection is upstream; investigation is downstream.

### What claims-automation trends should carriers watch in 2026?

Three stand out. First, straight-through processing of low-complexity claims is expanding; Crawford & Company expects more simple claims to auto-approve without an adjuster. Second, generative AI is moving from pilots to selective production - 78% of P&C insurers have adopted it, but only 4% have scaled it in claims, so the gap between experiment and operating model is the real story. Third, the un-automated stages are becoming the differentiators: with FNOL, estimating, and detection commoditizing, the investigation of flagged claims is the next loss-cost lever, since carriers investigate only about 25% of flagged claims today. Watch for autonomous investigation agents that lift that coverage toward 100%.

### Which vendors automate which part of the claims process?

Each layer has its own vendors. Claims management and FNOL workflow run on Guidewire ClaimCenter, Duck Creek Claims, and Snapsheet. Damage estimating runs on Tractable and Verisk's Xactimate. Fraud detection and scoring run on FRISS, Shift Technology, and Verisk ClaimSearch. These layers are well covered. The investigation layer - autonomous, end-to-end investigation of a flagged claim with an audit-ready report - has no named software incumbent; the current incumbent is the manual SIU team at every carrier. That is the layer Hesper AI is purpose-built for, sitting downstream of detection and upstream of the human SIU lead's adjudication decision.

### Does automating investigation replace SIU investigators?

No. The model shifts the investigator's role from execution to decision-making, not out of a job. Today an investigator spends 14+ days per case manually running collection, OSINT, database checks, and report assembly, and can only reach about 25% of flagged claims. An autonomous investigation agent runs those phases and produces a draft audit-ready report; the investigator reviews, overrides where needed, and signs the decision that goes into the antifraud-plan record under NAIC Model Act 680 and the documented-decision requirement of California 10 CCR 2698.36. SIU headcount gets re-aimed at higher-judgment work and at the 75% of flagged claims that previously went uninvestigated, rather than reduced.