Skip to main content
a68 Fraud Catcher for insurers

Revolutionizing
insurance fraud
detection.

Automatically identify suspicious claims and prioritize them for review.

a68 Fraud Catcher reviews documents, images, invoices, policy context, and claim history to help insurers reduce manual effort and prevent unjustified payouts.

a68 live
PDF
IMG
T&C
RISK
Repair invoice KS-2025-0816 submitted with the wildlife collision claim
Invoice validation
€5,414.50
Front-left vehicle damage photo submitted with a wildlife collision claim
Damage photo
Agent findings
12 flagged agents
Manual review recommended
Claim conflict
Image, invoice, and claim context show conflicting signals.
Photo plausibility check failed
Document scope appears inconsistent
Supervisor agent escalated the case
Live supervisor score
86%
suspicious
Scene mismatch and invoice anomaly detected
10%
Suspicious claims

Of all incoming claims show indicators of manipulation or fraud.*

* Source: Gesamtverband der Deutschen Versicherungswirtschaft (GDV) — „Versicherungsbetrug: Jede zehnte Schadenmeldung ist verdächtig.“

AI fraud scales
AI makes fraud scalable

What once required effort and technical expertise can now be created at scale in seconds: realistic claim photos, invoices, and supporting documents.

Review doesn't
Fraud scales. Manual review does not.

While manipulated images and forged documents can be created faster than ever, claims teams cannot forensically review every case by hand.

Competitive Advantage
Automation becomes a competitive advantage

Insurers that automate claims processes can settle claims faster, reduce pressure on review teams, and identify fraud signals more effectively.

a68 Fraud Catcher works like a digital SIU for every claim.

a68 Fraud Catcher orchestrates specialized AI agents, computer vision models, traditional machine learning methods, and forensic review modules across the entire claims process. This enables photos, documents, invoices, and case data to be systematically analyzed and consolidated into a clear, traceable assessment for claims teams.

Document Forensics

Surfaces tampered or altered documents that would slip past a manual review.

Content Analysis

Sanity-checks claim figures and counterparties against trusted external benchmarks.

Supervisor
Agent

Image Forensics

Flags synthetic, manipulated, or otherwise non-authentic imagery submitted as evidence.

Image Content

Checks whether the reported damage is plausible and whether the same evidence has appeared before.

A clear interface for complex claim decisions.

Walk through a live claim end-to-end — analyze the evidence, watch a68 Fraud Catcher reason, and review the final fraud assessment.

If the demo doesn't load, open it in a new tab.

Reliable fraud signals without the black box.

a68 Fraud Catcher reviews submitted evidence across documents, images, and case context to help insurers identify claims that deserve closer attention. Each assessment is explainable, auditable, and built for human review.

Scientific collaboration
Technical University of Munich ENS Paris-Saclay

Research-backed forensic AI

The forensic AI behind a68 Fraud Catcher is developed with leading research institutions, including the Technical University of Munich and Université Paris-Saclay, combining advanced AI agents with forensic methods for photo, document, and claims analysis.

GDPR-compliant by design

a68 Fraud Catcher is built for trusted insurance workflows with data protection, auditability, and controlled processing at the core.

EU privacy standards Auditable claim reviews Controlled data access

Evidence Consistency

Reviews whether documents, photos, and claim details tell the same story.

Document Integrity

Identifies unusual structures, duplicated content, and invoice-level inconsistencies.

Entity Validation

Checks whether companies, addresses, and commercial details appear plausible and consistent.

Image Credibility

Assesses whether submitted images fit the reported claim context.

Case Reasoning

Connects signals across files into one clear, explainable assessment.

Review Prioritization

Helps claims teams focus on cases that deserve manual attention first.

Proven with insurers

Established in live claims operations

a68 Fraud Catcher helps insurers reduce manual review, prioritize suspicious claims, and improve fraud detection across real claims workflows.

Bitkom — German digital industry association
Featured case study

Featured by Bitkom as an insurance AI case study

a68 Fraud Catcher was featured in Bitkom’s latest case studies, highlighting how AI can help insurers automate claims fraud detection, reduce manual review, and identify suspicious evidence earlier.

Read the Bitkom case study

Iris Tomingas

Head of ICT, INZMO

“a68 Fraud Catcher helped us strengthen fraud detection while making claims handling more efficient. Suspicious cases are reviewed more precisely, while unremarkable claims move faster through the process, increasing automation and relieving our claims teams.”

Read the full story

Markus Müller

CEO, Intec AG

“a68 Fraud Catcher relieves the claims department, accelerates payouts for valid claims, and sustainably lowers the claims cost ratio. This closes a critical gap in claims management.”

Read the full story
Business value

01 More automation 02 Less review effort 03 Fewer wrongful payouts

With a68 Fraud Catcher, insurers reduce manual review effort and prevent wrongful payouts with automated, scalable intelligence.

Higher efficiency in claims handling
Lower leakage and fewer wrongful payouts
Faster prioritization for SIU teams
YOUR BUSINESS CASE
Estimated gross ROI / year
€691,017
Gross benefit before licence, integration, and implementation costs.
Total benefit per year €691,017
Prevented wrongful payouts 77%
€540,000
270 Prevented payouts / year
Productivity gains 23%
€151,017
Handler hours freed
2,167 hrs
Released capacity for review, prioritization, and customer service
Prevented wrongful payouts + automation gain = annual ROI potential
Assumptions 27 prevented payouts per 1,000 claims 52% faster handling €69.7 fully loaded process cost / hr*

* Source: Destatis, labour costs 2025. For financial and insurance services, Destatis reports EUR 69.70 labour costs per hour worked.

Indicative figure based on conservative pilot assumptions. Actual net ROI depends on your claim mix and total cost of ownership.

Enterprise-ready integration.

a68 Fraud Catcher is API-first. It plugs directly into your existing core systems, including custom ERPs, for seamless, human-in-the-loop workflows.

a68 Fraud Catcher for your insurance lines.

Whether motor, property, liability, or supplemental health: a68 Fraud Catcher detects suspicious claim patterns, manipulated evidence, and implausible demands across lines.

Volkswagen Golf parked on a street representing motor liability and comprehensive insurance

Motor Liability & Comprehensive

Modern residential building representing buildings insurance

Buildings

Smartphone and consumer electronics representing property and electronics insurance

Property & Electronics

Family at home representing private liability insurance

Liability

Medical documents and stethoscope representing health and accident insurance

Private Health & Accident

Dog in everyday setting representing pet health and pet liability insurance

Pet Health & Liability

Aircraft wing above clouds representing travel insurance

Travel

Person reviewing contract documents representing legal expenses insurance

Legal Expenses

Your line is not listed?

Get in touch with us. In a free consultation, we will review together how a68 Fraud Catcher can support your fraud defense.

Request a free consultation

Detect suspicious claims with a68 Fraud Catcher before they become payouts.

Join forward-thinking insurers using a68 Fraud Catcher to protect their combined ratios.