Credit: Bartek/Adobe Stock
With the AI in insurance market set to exceed $10 billion in 2025, insurers are not only scrambling to utilize the technology, they’re working to use it better.
From risk mitigation and claims processing, to fraud detection and customer experience, the need for AI solutions is heightened attributed to a data explosion and demand for operational efficiency and cost reduction.
Recently, PropertyCasualty360.com spoke to Mike Cwynar, senior vice president of product delivery at Mitchell, an Enlyte company, about what AI is doing for claims management.
PropertyCasualty360.com: How does AI enhance claims prediction and triage?
Cwynar: AI's power in claims prediction comes from its ability to quickly analyze unstructured and/or historical data patterns that humans might miss. The technology can identify early indicators that a claim might need specialized attention, allowing adjusters to intervene sooner rather than later.
AI’s predictive analysis capabilities help create stronger, more individualized personalized care plans by incorporating best practice strategies and guidelines. So, if a case manager finds that an injured employee has certain comorbidities he’s living with, that information can be entered into predictive analysis programs to help determine how these diagnoses might also impact the person’s recovery time from injury.
With that information, personalized care plans can be created which address other issues aside from the initial injury to offer a more reliable timeline while reducing costly recovery delays.
PropertyCasualty360.com: Can you describe AI's virtual assistant capabilities?
Cwynar: A virtual assistant has the capabilities to quickly and efficiently gather all the information that's needed, sort through loads of structured and unstructured claims data and offer solid claims management recommendations.
These platforms can dramatically improve efficiency by handling routine tasks, allowing human adjusters to focus on more complex aspects requiring judgment and empathy. The technology also democratizes data exploration, allowing domain experts without technical backgrounds to extract meaningful insights.
Imagine being able to simply query, 'Give me the top ten procedures by provider in Pennsylvania for the last three years,' and receive immediate, actionable insights. There’s tremendous value in AI working behind the scenes as an assistant, but it can’t replace the human element of a competent claims adjuster.
PropertyCasualty360.com: How can AI improve medical summaries and documentation?
Cwynar: AI excels at synthesizing large volumes of medical information into concise, actionable summaries, but you still need that clinical expert to verify the information.
For instance, Enlyte’s Medical Records Summarization (MRS) solution for auto injury adjusters helps alleviate the burden adjusters often experience managing complex, disorganized documents that demand hours of analysis, potentially leading to missed details and delayed claims resolutions. The AI platform transforms lengthy medical documentation into targeted, adjuster-optimized summaries that prioritize clarity and relevance for expedited claims processing.
The output provides a clear, actionable summary aimed at accelerating claims evaluations and reducing oversight errors, while providing medical facts to promote decision-making. However, because the human element is so important, each claim that’s processed through MRS is verified by a clinical expert. We all know what can go wrong when AI outputs are blindly accepted, especially on high-exposure claims. So, it’s important to follow a trust-but-verify approach.
PropertyCasualty360.com: How does AI detect fraud and improve compliance?
Cwynar: Fraud detection represents one of AI's most promising applications in claims management. Pattern analysis has revealed billions being billed through specific providers using temporary procedure codes. These were insights that previously required extensive technical investigation.
The beauty is that as a domain expert, you can now explore data in ways you couldn't before. The ability to ask open-ended questions about billing patterns and then iteratively refine them represents a transformative approach to fraud detection.
However, compliance remains a critical concern, especially as regulations evolve. The National Association of Insurance Commissioners has developed governance guidelines for insurers that emphasize transparency, fairness, and accountability in AI development and deployment. Most states have adopted the model or at least portions of the framework and that’s especially important as AI continues to play a role in claims management.
PropertyCasualty360.com: What roles, if any, will AI begin to take from flesh and blood employees?
Cwynar: AI can do a lot of things, but it can’t replace the human element. Responsible use, especially in our industry, is vital. It means technology should be used behind the scenes for automation to help industry experts do their jobs more efficiently and better.
When it comes to adjusting claims or deciding treatment or giving care, there should always be a human in the loop. The most successful implementation strategy when it comes to AI is the "crawl, walk, run" approach. Organizations should start by clearly defining problems they aim to solve and testing technology against scenarios they already understand, viewing AI not as a complete solution but as a sophisticated assistant that remains subordinate to human expertise, critical thinking and compassion.
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.