Demystifying life actuarial technology: what you need to know about Generative AI

What is Generative AI?

Artificial intelligence (AI) is the emulation of human intelligence in machines that simulate learning, reasoning, problem-solving and decision-making as if they were human. Previously, AI tools were often designed to perform specific tasks, such as predictive modelling or classification. However, Generative AI (GenAI), a subset of AI, can create new content such as text, images and code. GenAI tools do this by identifying and replicating patterns found within their training data; they can synthesise information from large datasets and generate coherent and contextually relevant output in a human-like style. Ongoing advances in the technology are expanding GenAI’s ability to reason, understand nuances and to support complex decision making. This will make it increasingly relevant to domain-specific work.

What can AI do for actuaries?

The potential for AI to assist professionals in insurance and financial services is huge. Common use-cases include:

• Managing e-mail and diaries, taking minutes in meetings
• Assisting with typical Office applications, such as Word documents, Excel spreadsheets and PowerPoint presentations.
• Assisting in summarising information and drafting reports

However, all AI use cases require careful monitoring, both at deployment and in ongoing use, to avoid incorrect information or potentially discriminatory outcomes. A good approach is to think of these AI models as a highly capable junior colleague that produces great first drafts but still needs a human expert to review and expand upon the output.

How is AI used in life financial modelling?

WTW has proved a number of use cases, including:

• Explaining, writing, or debugging actuarial code
• Documenting models in natural language
• Generating reports on the outputs of reporting exercises

Systems capable of performing these tasks typically use a subset of GenAI known as large language models (LLMs), such as GPT-4 and Llama-4. These models excel at solving discrete and well-structured problems; however, they face practical limitations when tasked with more complex or nuanced challenges.

Recently, a new generation of “reasoning” LLMs, such as OpenAI’s O-series, has emerged. These significantly increase the potential of AI systems to handle complex tasks that were previously out of reach for standard models. These reasoning models enable AI to “think” for a longer period before arriving at a final solution, enabling them to undertake more demanding work. In future, they may be able to generate entire actuarial models from scratch given a detailed specification.

It's important to note that many regulators and professional bodies have been clear that the output of an AI should not be published directly without human review – and that they should be viewed as tools to assist humans, who remain fully accountable for the output

To learn more:

This summary was correct as of July 2025.

Agentic AI is covered in another article.

We actively monitor developments in technology and research how they may be applied to financial modelling in a sustainable way. Talk to us to find out more about how this technology can help your business.

Email WTW ICT Software Enquiries and we’ll connect you to our local experts.