On 10 March 2026 we ran a webinar called AI in Insurance Claims: Myths, Realities & The ExCo Gap.
We received many more questions than we could answer live, so our expert speakers have addressed them here. For more information please contact us or request a demo.
How can we use AI to support change management?
AI can accelerate change management by helping organizations embed new ways of working into everyday claims processes.
- AI provides decision support and transparency. When systems explain how they reach a recommendation, claims teams can understand and trust the outputs rather than seeing the technology as a “black box”.
- AI can capture expert knowledge. The reasoning used by experienced adjusters can be embedded into models, helping organizations standardize best practice and share expertise more widely across the claims function.
- AI creates continuous feedback loops. Claims handlers can review AI outputs, correct them where necessary, and feed those improvements back into the model. Over time this creates a collaborative relationship where the system becomes increasingly aligned with expert judgement.
In practice, the most successful implementations treat AI not simply as automation, but as a tool to help teams adapt, learn and scale new processes more quickly.
How long before you see measurable benefits from an AI claims pilot?
In most successful Sprout.ai implementations, measurable benefits appear within around 12 weeks of deployment in production.
The initial value typically comes from automating high-volume tasks such as document ingestion, claim triage, and policy coverage analysis.
However, the timeline depends on two main factors:
- Integration with core claims and policy systems
- Change management and user adoption
When these elements are addressed early, insurers begin to see improvements such as:
- Faster claims decision times
- Reduced manual processing
- Improved consistency in policy interpretation
AI initiatives should be measured against clear business KPIs, such as turnaround time, leakage reduction, or customer satisfaction, rather than simply measuring technology adoption.
How do you bring claims handlers onside with AI adoption, especially if they’re worried about job impact?
The most effective approach is to see AI as augmentation rather than replacement.
AI is particularly effective at removing repetitive work such as:
- document classification
- data extraction
- policy wording search
- initial claim triage
This allows claims handlers to focus on the areas where human expertise matters most: customer interaction, complex judgement, and negotiation.
Successful pilots usually do three things well:
Involve handlers early
Experienced handlers should be involved in pilots so they can shape how the system works.
Provide transparency and explainability
AI outputs should clearly show why a recommendation has been made so handlers can review or override it.
Create AI champions within claims teams
Early adopters inside the team can help drive adoption across the wider organization.
Train handlers in AI
Education and training can help handlers to develop increasingly important skills in the use of AI.
When handlers see AI removing administrative burden rather than replacing expertise, adoption tends to accelerate quickly.
What documentation or oversight do regulators expect when AI assists with claims decisions?
Regulators increasingly expect clear governance and traceability when AI supports claims decisions.
Key expectations typically include:
Decision traceability
Insurers should be able to demonstrate how an AI-supported decision was reached.
Audit logs
Every step in the decision process should be recorded and retrievable in case of regulatory review or customer complaints.
Human oversight
Even where AI assists decisions, accountability remains with the insurer or claims adjuster.
Data retention and documentation
Many insurers store AI decision logs for multiple years to align with regulatory expectations.
In practice, governance should be designed into the system architecture from the start, rather than retrofitted later.
What new skills are claims adjusters going to need in the rapidly developing claims AI workplace?
The claims role is evolving rather than disappearing. Future claims professionals will benefit from skills such as:
Data literacy
Understanding how AI tools analyze and interpret claims data.
AI collaboration
Working with AI outputs, validating recommendations, and providing feedback to improve models.
Broader claims expertise
AI can make knowledge more accessible, allowing adjusters to work across different claim types.
Empathetic customer management
Adaptability and continuous learning
Technology cycles are shortening, so the ability to adopt new tools quickly will become increasingly important.
Overall, the role is shifting from manual processing toward decision-making, judgement and customer outcomes.
What’s the best way to engage senior leaders in backing investment in claims AI?
The most effective way to secure executive backing is to focus on measurable business outcomes rather than technology.
Strong proposals typically highlight improvements such as:
- claims leakage reduction
- improved loss ratios
- faster claim settlement times
- improved customer experience
Claims leaders should also align AI initiatives with wider business strategy, showing how they support operational performance and competitiveness.
Executives respond best when they see:
- a clear return on investment
- defined KPIs
- a credible roadmap for scaling the solution across the organization
AI itself should never be the objective. The objective should always be better business outcomes.
How much of an issue are doctored real photos in fraudulent claims today?
Image manipulation is becoming an increasing challenge for insurers.
With the growing availability of generative AI and advanced editing tools, fraudsters can now:
- alter the severity of damage in images
- reuse photographs across multiple claims
- manipulate supporting evidence
Modern fraud detection approaches typically combine:
- image recognition models
- metadata analysis
- cross-claim comparisons
- anomaly detection techniques
Together, these tools help identify patterns or inconsistencies that may not be obvious to human reviewers.
What are the major obstacles and challenges insurers face in adopting AI?
Several recurring challenges continue to slow adoption.
Treating AI as a technology experiment rather than a business transformation
Projects succeed when they start with a clear operational objective.
Integration with legacy systems
Claims and policy systems can be complex and difficult to integrate with new technologies.
Scaling from proof-of-concept to production
Many pilots succeed technically but fail operationally due to governance issues, edge cases, or data challenges.
Organizational change and adoption
Even with executive sponsorship, success ultimately depends on adoption by claims teams. In most cases, the biggest barrier is not the technology itself – it is aligning people, processes, and systems around measurable business outcomes.
Follow or connect with the speakers on LinkedIn:
Ian Thompson, Global Insurance & Claims Expert, IMT Advisory