The Legacy Technology Problem: Why P&C Insurers Can’t Ignore Their Aging App Backlog

Most P&C insurers have modernized their core systems but continue to rely on a backlog of legacy applications built on aging IBM, Microsoft, and Java technologies, legacy collaboration platforms, and database-centric architectures. These are often difficult to maintain, poorly supported, and increasingly risky. Agentic modernization makes it faster, safer, and more cost-effective to elevate these applications to an AI-ready state. 

Who This Is For

CIOs, CDOs, and VPs of application development at midsized P&C insurance carriers ($500 million to $2 billion in revenue) who have an AI road map they believe in but a legacy estate that quietly undermines it. 

In Brief:

• The maintenance burden is real and growing. Nearly 75 percent of insurance carriers still rely on legacy technology, according to the Earnix 2024 Industry Trends report. The money spent keeping those systems running can’t fund the AI initiatives already on the road map. 

• Aging legacy applications represent a category most carriers haven’t addressed. These include traditional mainframe systems and the applications built and deployed during earlier efforts to move off the mainframe: VB6 and .NET Framework applications, Java/J2EE portals, SharePoint workflows, and client-server platforms such as PowerBuilder and Oracle Forms. 

• Many of these newer legacy applications run on aging or unsupported technologies. They’re typically poorly documented and deeply embedded in adjacent systems and business processes, making them difficult to modernize or replace. 

• Agentic capabilities are changing what’s economically viable to modernize. In our experience, rewrites that once took 12–16 weeks of manual engineering at unpredictable cost can now run in roughly two to four weeks at 30–50 percent lower cost. That changes the ROI case and makes the backlog solvable. 

Most P&C carriers have checked the core modernization box by upgrading policy administration systems, rebuilding claims platforms, or migrating core infrastructure. That work was hard and expensive, but it has largely been completed. As a result, many of the oldest technologies in application portfolios have been sunsetted. 

What most carriers haven’t solved is the long tail that came after: the VB6 portals built to replace something older, the Classic ASP claims intake screens, the antiquated workflows running in legacy document management systems that nobody has touched in years. 

In fact, according to the Earnix 2024 Industry Trends report, 74 percent of insurers still rely on legacy technology for core functions. And for most of them, the barrier isn’t the core platform, but the legacy estate surrounding it. 

This blog post makes the case for why the legacy problem deserves a board-level conversation and why the timing is better than it’s ever been. We’ll cover why the maintenance drain is a budget reallocation argument rather than a tech argument, why this category of applications creates a specific and underestimated AI readiness risk, and why agentic capabilities are changing what’s economically viable to fix. 

Why the Soft Market Makes Legacy Technology Modernization a Board-Level Cost Problem

The first challenge for upgrading legacy applications is that insurance premiums are softening. In the P&C sector especially, premiums softened through late 2025 and into 2026, and carriers are far more focused on operational efficiency than launching expensive transformation programs. In that environment, a recurring cost that drains the budget without adding value becomes a board-level problem. 

As premium growth slows and margins tighten across the property and casualty market, every line of spend is being scrutinized. Few lines are easier to question than money spent keeping decades-old applications alive. Usually, these systems do exactly what they did a decade ago, but they cost more each year to maintain. 

In a soft market, the legacy estate is no longer a tolerated cost and becomes a target. The carriers winning the efficiency conversation are the ones turning maintenance spend into capacity for what’s next. 

Application modernization that pays for itself in maintenance savings and frees capital for more transformative work becomes a margin argument instead of a technology request. Your CFO will appreciate the distinction. 

The Maintenance Burden: Where the Budget Actually Goes

Most insurers underestimate how much of their capacity is consumed simply by keeping the lights on. Clearwater Analytics research found that 93 percent of insurers say legacy technology is actively constraining their business. For a carrier with a heavy legacy estate, that constraint shows up first in the budget. 

These legacy applications are disproportionately expensive relative to the value they deliver. Maybe the vendor stopped supporting the framework years ago. Maybe the documentation was never written or is wrong. Or the developers who understood the code have retired or moved on, leaving a system that costs more to maintain each year while simply maintaining your current operations. 

However, every dollar you spend maintaining a legacy system that’s not adding value is a dollar unavailable for the AI, data, or digital road map. That direct trade-off is what turns a maintenance line item into a strategic constraint. 

The cost also compounds quietly. The same Clearwater research notes that the colleagues who know how these systems work are retiring, and finding people willing to maintain older platforms is a growing challenge, one that is particularly acute with newer legacy technologies. It’s easier to find a COBOL or RPG developer these days than somebody who can support Lotus Notes or PowerBuilder. 

As institutional knowledge walks out the door, the risk of a maintenance change breaking something grows, as so does the premium a carrier pays to the shrinking pool of people who still understand these systems. 

Why Legacy Technology Is the Hidden Blocker in Your AI Strategy

The AI readiness gap is real, and it’s widening. Gartner research found that only 20 percent of low-maturity organizations keep AI initiatives in production for any meaningful length of time. The gap between investment and production has many causes — adoption at scale, cost visibility, and regulatory concerns among them. But for insurers, fragmented legacy architecture is often the most concrete and fixable one. 

Insurers feel this acutely. The EIOPA Generative AI Market Survey, based on responses from 347 insurance carriers, found insurers scaling generative AI cautiously as they focused mostly on internal efficiency use cases that are held back, in part, by the systems the new tools must connect to. 

This is where the legacy estate quietly does the most damage. An AI solution can be production-ready and still fail to deploy because the application it needs to connect to can’t expose the integration layer the model requires. Each legacy application in a workflow becomes a discrete blocker for each AI initiative that touches it. 

The applications you’ve deprioritized for years may be the exact systems standing between your funded AI initiatives and production. That’s a different problem than a generic technical debt backlog, and it carries far more urgency. 

It also reframes the buying conversation across personas: 

• The CIO hears an AI strategy argument. 

• The VP of application development hears a throughput and execution argument. 

• The enterprise architect hears an architectural risk argument. 

The same legacy system sits at the center of all three. 

Why Agentic Capabilities Change Modernization Math

For years, these systems stayed on the backlog for one reason: manual rewrites were too slow, too expensive, and too risky to justify against the value the application delivered. Agentic modernization changes that math, making this the right moment to act on the backlog. 

For our clients, we’ve compressed multi-agent execution taking 12–16 weeks of manual rewrites to roughly two to four weeks at 30–50 percent lower cost. 

One reason for this acceleration is that specialized agents can greatly speed up requirements gathering. Agents reverse-engineer the existing application to surface hidden dependencies, undocumented business logic, and integration risks before a line of new code is written. For systems where the code is the only documentation, that capability eliminates the mid-project surprises that historically inflated scope and killed projects. 

None of this requires a transformation program. Instead, a legacy app rewrite can be approved as a discrete project with no board mandate, no yearslong engagement, and no disruption to operations while the work runs in parallel. Progress happens application by application, and the maintenance savings from each modernization project help fund the next. 

Start Using Agentic Modernization to Get Your Legacy Applications AI-Ready

The legacy problem has been a known cost for years at most P&C carriers, but it was tolerated because the alternative seemed too expensive. 

But now, the alternative has changed. The applications that stayed on the backlog because full rewrites weren’t viable may now be addressable in weeks at a cost that justifies itself against what carriers currently spend just to keep them running. 

The AI road map adds a layer the maintenance cost argument alone doesn’t carry. When a legacy application is blocking a funded AI initiative from production, the modernization decision is no longer about the application. It’s about the AI investment behind it. 

That’s a more manageable conversation than the program-level transformation mandate most teams have been avoiding. It starts with a specific system, a specific blocker, and a specific outcome. For carriers ready to look at their estate through that lens, the question worth asking isn’t whether to modernize, but which application is costing the most while blocking the most AI initiatives. 

Centric Consulting’s software engineering and development team works with insurance carriers to modernize legacy applications using agentic delivery, compressing multi-month rewrites to weeks at lower cost.

Frequently Asked Questions