For many years, digital transformation has been an elusive dream among insurers, brokers, and advisors. Organisations have deployed countless tools to re-engineer their systems to make them faster and more efficient but, if we are honest, most of these transformation efforts have missed the mark. There are three main challenges consistently impacting IT modernisation in insurance institutions. Firstly, the IT landscape is saturated with legacy technology which leaves data isolated in silos and creates crippling technical debt.
|
By Guy Moas, SVP R&D Life & Annuity, Sapiens
These existing systems slow progress but added to this problem are the standalone innovative applications, which often become a financial burden on top of the core architecture. They exacerbate system fragmentation and add redundant stages to multi-step processes.
The Core System Bottleneck
The third challenge is that most insurers operate across at least two core systems, but thanks to a steady stream of mergers and acquisitions (M&A) these systems have swelled over time, and have never been fully integrated. In my experience, I have observed organisations simultaneously running up to 16 distinct policy administration systems.
This complexity persists because the structural core of mature insurance organisations still heavily relies on decades-old COBOL and Assembler code sitting on legacy mainframes. It has a massive footprint as highlighted in a 2017 Reuters’ study that estimated roughly 43% of banking and insurance systems still run on COBOL. These "Big Iron" servers provide reliability, but they completely hamper an insurer's desire to move fast and launch agile digital products.
In the past, the IT department’s response to streamlining systems following M&A activity has been focused on desperate "cut-and-shut" or "lift-and-shift" integration attempts. This may create a sleek new front-end tool to let a company generate policies quickly but it forces integration of backend systems full of complex, custom-coded legacy infrastructure which inevitably holds insurers back. Unique operational lifecycles in life insurance complicate the process, such as the requirement that a policy may remain active for 60 years or more.
AI to the Rescue?
Digital transformation is a remedy that has frequently underdelivered, but now I believe that agentic AI has a unique capability to drive real, systemic change. What makes agentic AI truly exciting is its unprecedented ability to safely and intelligently port legacy code over to modern architectures.
Agentic AI can systematically read, map, and document ancient COBOL or Assembler frameworks, rather than just acting as a conversational chatbot. It has the potential to automate the heavy lifting of code translation, which dramatically reduces the timeframe for transformation. This ensures that what used to be a multi-year, high-risk manual process becomes an accelerated, controlled evolution.
Driving Real-World Value
This shift is no longer theoretical. Forward-looking market leaders are proving that insurers don't have to be paralysed by their historical infrastructure. They are automating the migration of their core systems to transform rigid code environments and streamline policy administration, which will enable them to deploy AI-driven capabilities at scale.
The winners in the insurance sector won't be those who watch from the sidelines, or those who settle for superficial digital patches. The future belongs to insurers who use modern AI infrastructure to liberate their core logic from legacy code and turn decades of institutional data into their ultimate competitive advantage.
|