Life and health insurers today face a complex balancing act. Product portfolios continue to grow in sophistication, regulations intensify, and financial reporting teams are under unprecedented pressure to produce fast, accurate, and explainable results. At the same time, operational costs and talent pressures remain stubbornly high.
Against this backdrop, artificial intelligence has rapidly moved from a talking point to a transformational force. AI is often described as “a fast, confident, tireless, occasionally wrong junior analyst” – and while that may raise a smile, it also reflects a reality: AI is powerful, imperfect, and increasingly expected to transform all workflows.
So, what is AI actually doing in actuarial and financial reporting today? And what might its emergence mean for the people and processes behind the numbers?
Go deeper with GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
How AI is helping to drive efficiency
Over the past decade, AI technologies have evolved from simple statistical models to advanced foundation models, reasoning engines and now AI “agents”. This progression is no longer theoretical; insurers are already applying standard tools to accelerate routine work and reduce operational burden, boosting team efficiency by about 30%. The AI agents are increasingly contributing to human efficiency, leveraging advanced software capabilities or functioning as actuarial accelerators to execute specific tasks as needed.
One area where this is particularly visible is model documentation and code translation. At WTW, for example, we have developed tooling that converts actuarial code into clear, natural‑language documentation, saving around 75% of the effort on much hated essential activities. And as part of a wider project, AI tools that translate open-source code and Excel spreadsheets, as well as building from specifications, can reduce the overall implementation costs by a similar proportion.
We see insurers similarly deploying AI to support data validation and cleansing; bulk document parsing; trend and variance analysis; narrative drafting and financial report preparation workflows; in addition to the more prevalent customer service triage.
Individually, these use cases offer incremental gains. But when stitched together – and especially when used within agentic architectures – the impact compounds quickly. Many insurers now see 20–30% efficiency improvements in reporting cycles where AI has been embedded purposefully.
AI vs humans: Augment or replace?
Fears that AI will replace actuarial or financial reporting talent are understandable but, for now, overstated. Judgment, accountability, regulatory interpretation and interpersonal communication remain fundamentally human responsibilities.
However, the nature of early‑career and mid‑career work is changing. Traditionally, analysts built expertise through repeated exposure to data preparation and production tasks. As AI increasingly replaces this work, entry‑level roles will shift rapidly toward interpretation, scenario analysis and communication of results.
This transition brings three major consequences:
Organisational design will change: Continuing the trend seen with automation in recent years, teams built around large production functions will shrink. Fewer people will be needed to generate numbers; more will be needed to challenge, interpret, narrate and govern them.
Skills portfolios must expand: AI literacy will become as fundamental as spreadsheet literacy once was. Those who thrive will be those who can use AI as a collaborator rather than a novelty tool.
Recruitment patterns will shift: Graduate hiring pipelines may narrow in the short term as automation removes the need for large analyst cohorts. Yet regulators retain their requirements for responsibility, with a strong onus on senior management for validation of AI‑assisted outputs. This will drive a new generation of graduates who learn their trade through challenging rather than doing.
Crucially, the greatest barrier for most teams today is not technology, it is thinking too small. Asking AI simply to fix known errors in a dataset misses the opportunity to validate the data for unexpected issues or even to redesign the end‑to‑end process. Creativity and vision will differentiate the winners from the followers.
Human oversight still matters in AI
Even though actuarial work rarely involves personal data with its associated bias and confidentiality risks, financial reporting sits within one of the most tightly regulated environments in the corporate world. Model governance frameworks, audit trails and sign‑off processes leave little room for opaque or unexplained AI behaviour.
Therefore, AI outputs must be reviewed by accountable humans; controls must evolve to include prompt governance, explanation frameworks, and AI‑specific testing; and corporate governance teams should be partners, not gatekeepers.
Modern AI systems can already perform coding tasks, run high‑level checks, generate draft commentary and review processes for operational weaknesses. But they must operate under human supervision, much like training and checking the work of a new colleague.
The rise of agentic AI
The next leap in capability comes from agentic AI – systems that can plan tasks, execute multi‑step workflows, interact with IT systems and use tools semi-autonomously.
For financial reporting teams, this is achieved through a combination of flexible AI questions combined with robust models and reporting processes. Together it could mean real‑time answers to “what if?” questions; dashboards that update themselves when the market moves; automated change testing and reconciliation workflows; and reduced delays from specialist technical teams
However, these gains bring governance challenges. Today’s AI often looks “magical,” which can undermine trust. One promising mitigation is neurosymbolic AI, combining machine‑learning‑based pattern recognition with explicit rules – making outputs more explainable and auditable.
AI is moving from passive assistant to active co‑worker. It has the potential to add huge value in complementing humans. The question for insurers is no longer if this technology will transform reporting, but how quickly they can adapt.
Mark Brown, Global Proposition Lead, Life Financial Modelling Insurance Consulting and Technology, WTW
