Read our article published by InsuranceERM
Agentic artificial intelligence (AI) will increasingly impact actuarial modelling and insurers, according to Martin Sher, chief operating officer of UK-headquartered actuarial and technology consultancy, Zenith Actuarial.
Agentic AI refers to AI systems and models that can act autonomously to achieve goals without the need for constant human guidance. To achieve this level of autonomous decision-making and action, agentic AI relies on a complex ensemble of different machine learning, natural language processing and automation technologies.
Discussing how automation in actuarial modelling will evolve, and the impact of AI, Sher said agentic AI is already gaining traction and this is likely to continue.
He said: “The more these models are given agency, where they can ‘think and do’ by themselves and interact with other EUC [end-user computing] processes, the more the need to ensure a robust governance structure is in place.”
Sher stressed that insurers “should now be thinking about how they implement governance, control and automation solutions and policies” given the rise of agentic AI.
“This approach should not stifle innovation, but equally should also avoid the same EUC governance, automation, scalability and key person risks as spreadsheet usage does,” Sher added.
Sher described the widespread adoption of AI/Gen AI within actuarial teams as “inevitable”, even if the specifics of how and for are yet to be determined.
“For example, we have developed an agentic AI solution that, given a data file and a specified actuarial objective (e.g. “fit a … model to …”), iteratively solves for the solution. We also use AI to automate data transformations from source data to analytic model data schemas, and to automate data science pipelines.”
Zenith Actuarial won Solvency II solution of the year award for its evo-insight solution in InsuranceERM’s 2024 UK & Europe awards.
The actuarial and technology consultancy also features in InsuranceERM’s technology guide 2024/25.
Article published by Ronan McCaughey at InsuranceERM - 17 April 2025.
Insights
