What can AI automate for a Actuary?

AI can help with: Drafting first versions of technical memos, assumption logs, and review documentation; Writing and debugging code for reserving, pricing, and capital models; Cleaning, reconciling, and reshaping large claims and policy datasets; Summarizing regulatory updates and lengthy technical reports; Building routine dashboards and recurring exhibit calculations.

What stays distinctly human for a Actuary?

Still human: Setting and owning the professional judgment behind reserving and pricing assumptions; Signing off on results and taking responsibility under actuarial standards; Judging when a model is inappropriate for the business context; Communicating uncertainty and trade-offs honestly to boards and regulators; Weighing ethical and fairness implications of rating and underwriting decisions.

AI for Actuary: tools, prompts, and how the role is changing

The shift

How AI is changing the Actuary role

In 2026, AI is handling more of the repetitive work in actuarial practice, from cleaning and reconciling large claims datasets to drafting documentation for reserving and pricing reviews. Language models help summarize regulatory updates and explain model assumptions to non-technical stakeholders, while coding assistants speed up the building and testing of pricing and capital models. The actuary's role is shifting toward reviewing AI output, validating assumptions, and owning the professional judgment behind the numbers.

What AI can take off your plate

Drafting first versions of technical memos, assumption logs, and review documentation
Writing and debugging code for reserving, pricing, and capital models
Cleaning, reconciling, and reshaping large claims and policy datasets
Summarizing regulatory updates and lengthy technical reports
Building routine dashboards and recurring exhibit calculations

What stays distinctly human

Setting and owning the professional judgment behind reserving and pricing assumptions
Signing off on results and taking responsibility under actuarial standards
Judging when a model is inappropriate for the business context
Communicating uncertainty and trade-offs honestly to boards and regulators
Weighing ethical and fairness implications of rating and underwriting decisions

Tools

Five AI tools for Actuarys

GitHub Copilot

An actuary uses it to write and debug R or Python code for reserving triangles, pricing models, and data pipelines faster.

ChatGPT

An actuary uses it to draft technical memos, summarize regulatory documents, and explain assumptions in plain language for committees.

Microsoft Copilot in Excel

An actuary uses it to build formulas, pivot summaries, and quick analyses across large policy and claims spreadsheets.

Claude

An actuary uses it to review long actuarial reports or solvency regulations and extract the relevant requirements and changes.

DataRobot

An actuary uses it to build and compare predictive models for lapse, mortality, or claims frequency without coding each one by hand.

Prompts

Five prompts to try today

Paste these into Claude or ChatGPT and replace the bracketed parts with your own details.

1. Explain a reserving method

Explain the [chain ladder / Bornhuetter-Ferguson] reserving method to a non-actuarial audit committee. Cover the key assumptions, when it breaks down, and what limitations I should flag for [line of business].

2. Draft assumption documentation

Draft documentation for the following actuarial assumption: [assumption description]. Include the rationale, data sources used, sensitivity to change, and a section on key uncertainties. Keep the tone formal and suitable for a peer review.

3. Code a pricing calculation

Write [R / Python] code to calculate a risk premium for [product] using these rating factors: [list factors]. Include comments, handle missing values, and add basic validation checks on the output.

4. Summarize a regulation

Summarize the key requirements of [regulation or standard, e.g. IFRS 17 / Solvency II] that affect [reserving / capital / disclosure]. List the practical actions an actuarial team must take and any recent changes from prior versions.

5. Sense-check model results

Here are my model outputs for [metric] across [periods or segments]: [paste data]. Identify any results that look inconsistent or unexpected, suggest possible causes, and list checks I should run before signing off.

A day in your inbox

This is the kind of brief a Actuary gets, every weekday morning.

The Morning Current

Weekday morning

✦ Personalized for: Actuary

Today's Tool

Using Copilot for a reserving script

GitHub Copilot can generate a first-pass R script to build development triangles and apply chain ladder factors from your claims extract. You still review every factor and tail assumption before trusting the output.

Today's Prompt

Sense-check unexpected reserve movement

Paste your reserve estimates by accident year into the model and ask it to flag any movements that look inconsistent with prior periods. Use its list as a starting point for your own investigation, not a conclusion.

Today's Trick

Make the AI show its assumptions

Ask the assistant to list every assumption it made before giving an answer on a calculation or method. This surfaces hidden errors and helps you catch where its reasoning diverges from your data.

AI for Actuarys