AI for finance and actuarial business processes

Tom Peplow, Principal, Milliman.

The impact of AI

Our imagination is our competitive advantage against artificial intelligence (AI). And it is that imagination that will let us solve society’s biggest problems. The problem I have chosen to focus my imagination on is improving retirement outcomes for all, including helping individuals provide for their loved ones in the event they can no longer work. Isaac Newton said, “If I have seen further it is by standing on the shoulders of giants.” I believe by using machines to automate previously manual processes, we will generate more insight from the data those processes will capture, and we will free up the creative energy of the people in the life insurance industry, making it possible for them to reimagine the insurance that I believe is a fundamental pillar of society.

There are significant operational benefits for end-to-end automation of business processes. Robotic process automation has been getting a lot of traction, driven by significant enhancements of artificial intelligence through the collection and leverage of vast amounts of data. One of the examples of this advancement that gets closest to human intelligence is Google Duplex (an AI system for accomplishing real-world tasks over the phone), made famous by Sundar Pichai’s demo of Google’s virtual assistant booking a haircut and then a table at a restaurant.

I meet a lot of actuaries in my role and I often confront a lot of cynicism around AI, especially how it can help reduce cost and risk. However, I believe AI offers huge opportunities to exploit data in innovative ways to reduce risk and create products better suited to the modern age. In automobile insurance, we are seeing how connected devices can be used to lower the cost of insurance to younger drivers, by identifying and rewarding good drivers. In health, we are seeing the rise of predictive analytics to deliver better care. In underwriting, we benefit from solutions that automate costly processes and leverage broader data sources to remove hurdles in customer acquisition while also reducing the risk exposure to the insurer.

As the technology giants provide the cognitive services (such as speech recognition) there will be few barriers to complex AI techniques, demonstrated by services like Google Assistant. Cognitive capabilities at the intelligent edge backed by the intelligent cloud will transform business processes. For example, data collected by cars (at the intelligent edge) is used to train models (in the intelligent cloud) to help prevent accidents. There will be a significant step change in how humans interact with machines and that will afford us more opportunity to innovate and create.

The opportunity for insurers

AI is an opportunity to reduce cost and risk in core business processes, but data is needed to enable the optimization. The first step on this journey is the organizational vision and will to pursue the modern data warehouse, in spite of the shadow of previous data warehouse failures. In the world of insurance, looming regulatory changes—such as International Financial Reporting Standard (IFRS) 17 and GAAP Long-Duration Targeted Improvements (LDTI)—require new data pipelines. Administration, actuarial, and finance functions need to be able to share data securely and robustly, with a complete audit trail (though Sarbanes-Oxley also required this). This is the perfect opportunity to start the journey to the modern data warehouse—to foster a culture of organization-wide insight and data sharing using modern technologies and to lay the foundation for AI and other advancements. The regulatory deadlines are necessitating action to achieve compliance. The opportunity is to use these regulatory drivers to also establish a new data foundation for competitive advantage through a modern data warehouse.

My next blog post will talk about how data is the new oil that powers the AI revolution.