There is a quiet renaissance underway in the actuarial profession, and most of us are too busy reconciling cashflows to notice.
For the better part of three decades, the daily work of being an actuary has been steadily reshaped by forces that have almost nothing to do with the original purpose of the profession. Solvency regimes got tighter. Reporting standards got denser. Data volumes exploded. Systems multiplied. Each well-intentioned addition pulled actuarial teams a little further away from the thing they were trained, qualified, and credentialed to do: bring professional judgment to bear on long-horizon uncertainty in a way that helps organisations make better decisions.
The good news is that the tide is turning. The better news is that the actuaries who recognise what is happening early will define the next chapter of the profession.
How we got here
Ask any actuary what consumes their working week, and the honest answer is rarely “thinking about risk.” It is data extraction, reconciliation, model runs, control documentation, audit responses, version comparisons, and explaining month-on-month movements. The actual actuarial part — the bit that requires the qualification — often happens in the margins, late in the cycle, under time pressure.
This is not anecdote; the numbers back it up. An Oliver Wyman survey of US insurers, summarised by the Society of Actuaries, found that actuaries at most insurance companies spend up to half of their time on manual work surrounding the pre-model, core model, and post-model environments. Research summarised by hyperexponential, drawing on Casualty Actuarial Society E-Forum work, puts the pre-automation split more starkly: roughly 70% of actuarial time on data manipulation and calculation, with only 30% on interpretation.
IFRS 17 deserves a particular mention here, because the experience of the past few years has crystallised the problem. As Baker Tilly has noted, it is the most significant accounting change for the insurance industry in at least twenty years, and the transition has demanded extraordinary actuarial bandwidth. EY observes that the standard’s data-intensive nature has made access to granular data effort-intensive, with insurers needing to develop or adapt actuarial models, implement new software, and expand financial reporting processes. The Actuary Magazine has documented that life and health carriers have experienced extended timelines across most financial reporting processes, with the workload heavy, the difficulty high, and the fault tolerance low.
Solvency frameworks, internal model approval, ORSAs, embedded value reporting, with-profits governance — each new layer adds modelling, documentation, validation, and sign-off. The cumulative effect is a profession in which sophisticated, expensively-trained professionals spend an increasing share of their time on activities a senior actuary from 1990 would barely recognise as actuarial work.
What actuaries were originally for
It is worth remembering where this started. When Edmond Halley published his analysis of the Breslau mortality tables in 1693, when James Dodson developed the level premium system, when the Institute of Actuaries was founded in London in 1848, the purpose was never data engineering. It was the application of mathematical reasoning and professional judgment to long-tail financial promises — promises that would only be kept or broken decades into the future, in conditions no one alive could fully foresee.
The Actuarial Control Cycle, taught to every qualifying actuary, captures this beautifully. As described in the British Actuarial Journal and reinforced across the syllabi of ASSA, the SOA, and the IFoA, it is a systematic approach to problem-solving in three steps: define the problem, design the solution, and monitor the results, all within a context shaped by professionalism, regulation, and the wider commercial environment. It is fundamentally a thinking framework, not a processing framework. Its power lies in the actuary’s judgment about which assumptions matter, what the answer actually means, and what decision should follow.
ASSA’s own definition is worth quoting in spirit, if not verbatim: an actuary applies analytical, statistical, mathematical, and strategic thinking skills to bring greater understanding and improve decision-making about uncertain future events. The strategic thinking part is doing more work in that sentence than the industry has lately allowed.
A revealing observation appeared in the Society of Actuaries’ own discussion of principles in the profession: standards of practice can become procedural, providing idealised lists of considerations and methodologies that may or may not be relevant in a given situation, without offering insight into the relative importance of considerations or how techniques actually solve the underlying problem. That gap — between procedure and understanding — is where actuarial judgment is supposed to live. It is also, increasingly, the gap that gets skipped when there is not time.
The asymmetry that matters
Larger insurers have, to a degree, been able to protect this judgment-driven role at the senior level. A Chief Actuary, a Head of Capital, a With-Profits Actuary, a Statutory Actuary — these positions exist precisely to give actuarial judgment a strategic seat. They tend to spend more of their time on interpretation, board engagement, capital allocation, and the kind of firefighting that comes with steering a complex balance sheet through volatile conditions. As Deloitte has put it, the modern Chief Actuary is increasingly viewed as a broader business leader, charting strategic direction and bringing true business value. The grunt work happens further down the hierarchy, but the thinking still happens at the top.
Smaller insurers and consultancies have not always had that luxury. A statutory actuary at a smaller life office, or a senior consultant at a boutique firm, often plays every role at once: data engineer, modeller, reviewer, communicator, advisor. The result, in practice, is that the most senior person gets pulled down into the technical detail because there is no one else to do it — and the strategic conversation never quite happens. The value-add that the textbooks describe gets squeezed out by the next reporting deadline.
This is the structural reality I want to name plainly, because it is the thing that AI and automation are about to change.
The inflection point
For the last few years, the consulting world has been writing about AI and actuaries with a tone that oscillates between cheerleading and reassurance. The honest assessment, stripped of marketing language, is this: a great deal of the work that consumes actuarial time today is exactly the work that AI is best positioned to take over.
PwC has described how generative AI in reserving could automate data extraction, produce quality reports highlighting discrepancies for actuaries to review, generate exhibits summarising trends, fit neural network models across lines of business, and automatically build reports and presentations summarising the results. RSM has documented how generative AI can draft year-end reserve adequacy memos by comparing current to prior results, explaining notable changes, and ensuring required sections are covered, with actuaries acting as final reviewers rather than initial drafters. Deloitte frames the broader shift as a move from operational efficiency to strategic value, with actuaries positioned as strategic advisers who translate complexity into action.
The numbers are starting to land too. McKinsey has estimated that artificial intelligence could generate as much as $1.1 trillion in annual value for the global insurance industry. McKinsey’s own work with Aviva, examined in industry coverage, shows what this looks like in practice: a portfolio of more than 80 AI models delivering substantial savings and cutting liability assessment time by weeks — with human roles shifting toward oversight, complex judgment, and strategic interpretation rather than being eliminated.
The post-automation time-allocation picture, in the same body of research, is the mirror image of where actuaries are today. Where teams used to spend 70% of their time on data manipulation and 30% on interpretation, the after-AI picture flips that to 30% on validation and monitoring and 70% on strategic interpretation and stakeholder communication. That is not a marginal shift. That is a different job description.
The Society of Actuaries’ commentary on agentic AI describes a near-future in which a digital assistant prepares data for a valuation run, identifies inconsistencies, suggests corrections, executes test suites overnight, flags anomalies, and drafts summary reports before the team logs in the next morning. None of that is science fiction; early versions are already being piloted in actuarial transformation programmes across insurers and consultancies. Legal & General’s own careers blog frames the question crisply: actuaries are not being replaced by AI; the role is changing, and the future belongs to those who can combine technical expertise with human insight.
What this means for the profession
If we take this seriously, the implication is striking. The activities that have crowded out actuarial thinking for thirty years — the cleansing, the reconciling, the documenting, the drafting — are precisely the activities most amenable to automation. The activities that have been squeezed — judgment, interpretation, strategic advice, professional contemplation, communication of uncertainty to decision-makers — are precisely the activities that remain durably human.
The opportunity is to use this moment to return to the original vision of the profession. Not to abandon technical rigour, but to free it. Not to do less work, but to do more of the right work. Not to compete with AI on speed of computation, but to ensure that what gets computed is the right thing, that the assumptions chosen reflect genuine professional judgment, that the results are interpreted properly, and that the business decisions they support are sound.
The Actuarial Society of South Africa’s leadership has been articulating something similar from a public-interest angle. ASSA’s view, captured in recent commentary by President Lusani Mulaudzi and his predecessor, is that the profession must be more than expert in risk; it must take a public-interest stance on issues like the two-pot retirement system, the National Health Insurance debate and climate change. That ambition is impossible if the profession is permanently buried in spreadsheet reconciliation. It becomes plausible only if the technical floor is automated and the judgment ceiling is raised.
A particular moment for young actuaries
Younger actuaries are in an unusually fortunate position. They have not yet built careers around manual processes that AI is about to disrupt. They are native to the tools that will reshape the work. And they are entering the profession at exactly the moment when the bar for what counts as actuarial value-add is being reset upward.
The temptation, for any young actuary watching this shift, is to lean further into tooling — to become the best Python user, the best modeller, the best automator on the team. That is useful but insufficient. The deeper move is to lean into what tooling cannot replace: the application of actuarial principles to problems that have not been solved before. The Actuarial Roles of the Future commentary in The Actuary Magazine argues for apprenticeship-style learning analogous to medicine and law, with junior actuaries getting structured exposure to board-level preparation, strategic decision-making, and the messy realities of how technical recommendations become commercial action. That kind of development is more important now, not less.
The qualifying syllabus, with all its emphasis on the Control Cycle, professionalism, judgment under uncertainty, and the wider commercial context, was always pointing at this. The frustration of the last two decades has been that the working life of a junior actuary often had very little to do with what the qualification was for. The next decade can change that.
For young South African actuaries in particular, the moment carries an extra dimension. There are still only around 1,500 qualified actuaries in the country, against an economy whose insurance, retirement, healthcare, and climate-resilience challenges are enormous. The profession’s relevance to those challenges has nothing to do with how quickly anyone can clean a data extract. It has everything to do with how well actuaries can think — and how clearly they can communicate that thinking to people who need to make decisions.
What needs to happen
A few things must be true for the profession to convert this opportunity into a genuine shift.
First, leadership in actuarial functions has to consciously redirect time saved by automation toward higher-order work, rather than allowing the gains to be absorbed by yet more reporting granularity. There is a real risk, in any transformation, that capacity freed up by technology is immediately re-deployed into more of the same.
Second, the profession needs to invest in the skills that are durable: communication, commercial fluency, ethical judgment, strategic framing, the ability to sit in a room with non-actuarial leaders and translate uncertainty into action. The Institute and Faculty of Actuaries, ASSA, the SOA, and the CAS have all signalled that the future actuary is more of a translator, a governance partner, an AI overseer, and a strategic adviser. The training, mentoring, and career structures have to catch up with that signal.
Third, smaller insurers and consultancies need to recognise that automation is, on balance, more of a leveller for them than for the large players. The same generative tools that an in-house team of fifty can deploy are increasingly available to a team of five. A boutique actuarial consultancy in Johannesburg can plausibly offer the same calibre of insight, on the same cadence, as a large advisory firm — provided it is willing to rebuild its workflows around judgment rather than throughput.
Fourth, the broader insurance industry needs to start treating actuaries as the strategic resource they were always meant to be. The conversations that matter — capital allocation, product strategy, climate exposure, demographic shifts, the long-run economics of new distribution models — are exactly the conversations actuarial training prepares people for. Leaving them to be conducted without serious actuarial input has been a quiet cost the industry has carried for years.
Back to the vision
The founders of the profession would, I suspect, be both proud and bemused by where actuaries find themselves in 2026. Proud, because the discipline they shaped has become indispensable to almost every long-horizon financial promise made in modern economies. Bemused, because so much of the daily work of their successors has so little to do with the kind of judgment they spent their careers cultivating.
The next few years offer the chance to put that right. The technical floor is being automated. The judgment ceiling is rising. The young actuaries who recognise this and lean into it — who treat AI as something that compounds actuarial thinking rather than threatens it — will find themselves doing exactly the work the profession was originally invented to do.
It is, in that sense, a renaissance. And it could not have come at a better time.
This article describes the engineering work our Finance Modernisation practice does — modernising the operating layer around actuarial models so the actuary’s time can return to judgment, not reconciliation. For the practical AI side, see our companion piece on working with AI agents.