Embracing the Future: Strategies for an AI-Driven World
Artificial intelligence (AI) is rapidly becoming essential for insurers aiming to remain competitive amid geopolitical volatility, evolving claims patterns, persistent cost pressures, and increasing customer expectations. To fully harness AI’s potential, carriers must first address shifting workforce dynamics, modernize outdated legacy systems, and tackle their tech-scoping challenges, as discussed by panelists during a recent webinar from Federato.
Unlike other technological advancements such as cloud computing or mobile applications, which primarily transform specific areas, AI necessitates a fundamental rethinking of business models and the creation of new opportunities across the entire value chain. Mark Breading, partner at ReSource Pro, likened this shift to the Internet’s emergence over 30 years ago.
AI Can Distribute Expertise
The insurance sector is currently experiencing a generational transition, according to Federato CEO and Co-founder Will Ross. As senior executives retire, there is a turnover in decision-makers, leading to newer employees being tasked with complex underwriting responsibilities they may not be fully prepared for.
Karlyn Carnahan, head of Insurance, North America at Celent, echoed this sentiment, noting that many basic tasks previously assigned to trainee underwriters have been automated, leaving only the more complex decision-making processes. “With a new workforce lacking 20 years of expertise, the challenge is ensuring the right decisions are made. This is where AI can significantly augment the process,” she stated.
AI can guide newer underwriters, but it must be built on the organization’s existing expertise. Carnahan emphasized the need for skill sets that differ from traditional roles, highlighting the importance of designing prompts for AI systems to break down tasks into manageable steps. This shift means that employees will increasingly supervise AI-driven workflows, which can dramatically accelerate decision cycles by automating tasks like drafting policies.
“Knowledge becomes more distributed,” she added, reducing reliance on individual expertise and allowing AI to learn from collective experiences. This capability can help carriers avoid repeating historical patterns of losing underwriting discipline during soft market cycles.
Typically, insurers shed business during hard market cycles, but as markets soften, they often revert to less disciplined practices. AI can help mitigate this by aggregating vast amounts of data, ensuring underwriting decisions are not solely dependent on individual knowledge.
Legacy Infrastructure is a Barrier
Panelists identified legacy systems as a significant barrier to leveraging AI’s full potential. These older systems often function as data silos, leading to fragmentation and duplication of efforts. While they perform basic functions like policy manufacturing and claims processing, they fall short of modern underwriting needs.
Carnahan pointed out that legacy systems are challenging to configure, making updates time-consuming and costly. Breading concurred, noting that a significant portion of IT budgets is consumed by maintaining these outdated systems, hindering innovation and market responsiveness.
Carriers often layer new systems on top of old ones, complicating integration. “The cost to sunset these systems is often hard to justify, leading to a burden of multiple policy administration and claims systems,” Carnahan explained.
Breading noted that many legacy systems were designed with a conservative mindset, focusing on sequential processing and static requirements, which do not align with today’s dynamic environment. The incremental modernization efforts over the past few decades have not sufficiently addressed these challenges.
Ross highlighted that simply adding AI interfaces to legacy frameworks is impractical due to their slow and fragmented nature. The time required for AI to retrieve data can be excessive, making it difficult to meet the instant results consumers expect.
Carriers Have a Tech-Scoping Problem
Carnahan observed that while AI has shown promise, many carriers struggle with project scope. Some focus too narrowly, limiting returns, while others aim too broadly, making projects unmanageable. “Carriers need to define their objectives carefully, ensuring use cases are significant enough to matter but not so expansive that they become unfeasible,” she advised.
Ross echoed this sentiment, advocating for a balanced approach to project scoping that considers organizational impact. “You need to provide answers regarding future capabilities and the scale of impact,” he said.
Near-Term Priorities
Looking ahead, Ross emphasized the need for carriers to modernize product definitions to expedite quoting processes. While the bind-and-issuance process can remain slower, the quoting process must be swift, necessitating access to rating and forms attachment logic.
Carnahan agreed, stressing the importance of understanding workflows and processes. “Without a clear understanding of current operations, envisioning future AI applications becomes challenging,” she concluded.
Artificial intelligence (AI) is rapidly becoming essential for insurers aiming to remain competitive amid geopolitical volatility, evolving claims patterns, persistent cost pressures, and increasing customer expectations. To fully harness AI’s potential, carriers must first address shifting workforce dynamics, modernize outdated legacy systems, and tackle their tech-scoping challenges, as discussed by panelists during a recent webinar from Federato.
Unlike other technological advancements such as cloud computing or mobile applications, which primarily transform specific areas, AI necessitates a fundamental rethinking of business models and the creation of new opportunities across the entire value chain. Mark Breading, partner at ReSource Pro, likened this shift to the Internet’s emergence over 30 years ago.
AI Can Distribute Expertise
The insurance sector is currently experiencing a generational transition, according to Federato CEO and Co-founder Will Ross. As senior executives retire, there is a turnover in decision-makers, leading to newer employees being tasked with complex underwriting responsibilities they may not be fully prepared for.
Karlyn Carnahan, head of Insurance, North America at Celent, echoed this sentiment, noting that many basic tasks previously assigned to trainee underwriters have been automated, leaving only the more complex decision-making processes. “With a new workforce lacking 20 years of expertise, the challenge is ensuring the right decisions are made. This is where AI can significantly augment the process,” she stated.
AI can guide newer underwriters, but it must be built on the organization’s existing expertise. Carnahan emphasized the need for skill sets that differ from traditional roles, highlighting the importance of designing prompts for AI systems to break down tasks into manageable steps. This shift means that employees will increasingly supervise AI-driven workflows, which can dramatically accelerate decision cycles by automating tasks like drafting policies.
“Knowledge becomes more distributed,” she added, reducing reliance on individual expertise and allowing AI to learn from collective experiences. This capability can help carriers avoid repeating historical patterns of losing underwriting discipline during soft market cycles.
Typically, insurers shed business during hard market cycles, but as markets soften, they often revert to less disciplined practices. AI can help mitigate this by aggregating vast amounts of data, ensuring underwriting decisions are not solely dependent on individual knowledge.
Legacy Infrastructure is a Barrier
Panelists identified legacy systems as a significant barrier to leveraging AI’s full potential. These older systems often function as data silos, leading to fragmentation and duplication of efforts. While they perform basic functions like policy manufacturing and claims processing, they fall short of modern underwriting needs.
Carnahan pointed out that legacy systems are challenging to configure, making updates time-consuming and costly. Breading concurred, noting that a significant portion of IT budgets is consumed by maintaining these outdated systems, hindering innovation and market responsiveness.
Carriers often layer new systems on top of old ones, complicating integration. “The cost to sunset these systems is often hard to justify, leading to a burden of multiple policy administration and claims systems,” Carnahan explained.
Breading noted that many legacy systems were designed with a conservative mindset, focusing on sequential processing and static requirements, which do not align with today’s dynamic environment. The incremental modernization efforts over the past few decades have not sufficiently addressed these challenges.
Ross highlighted that simply adding AI interfaces to legacy frameworks is impractical due to their slow and fragmented nature. The time required for AI to retrieve data can be excessive, making it difficult to meet the instant results consumers expect.
Carriers Have a Tech-Scoping Problem
Carnahan observed that while AI has shown promise, many carriers struggle with project scope. Some focus too narrowly, limiting returns, while others aim too broadly, making projects unmanageable. “Carriers need to define their objectives carefully, ensuring use cases are significant enough to matter but not so expansive that they become unfeasible,” she advised.
Ross echoed this sentiment, advocating for a balanced approach to project scoping that considers organizational impact. “You need to provide answers regarding future capabilities and the scale of impact,” he said.
Near-Term Priorities
Looking ahead, Ross emphasized the need for carriers to modernize product definitions to expedite quoting processes. While the bind-and-issuance process can remain slower, the quoting process must be swift, necessitating access to rating and forms attachment logic.
Carnahan agreed, stressing the importance of understanding workflows and processes. “Without a clear understanding of current operations, envisioning future AI applications becomes challenging,” she concluded.
