Why AI and regulation are hollowing the talent pipeline in banking, wealth and advice, and how to rebuild it
AI is dissolving the entry roles that trained the next generation of bankers and advisers. Regulation is raising the bar on the senior judgement that remains. The talent pipeline is being hollowed from both ends at once. Here is how to rebuild it, deliberately, before the gap becomes structural.
For people, risk & business leaders · Aotearoa New Zealand & Australia · 15 minute read
THE SHORT VERSION
Banking, wealth and advice is not facing a churn problem. It is facing a capability problem, and it is being squeezed from two directions at once. From below, AI is automating the routine, entry-level and processing work that was the sector's training ground. From above, conduct and competency regulation has raised the bar on the advice, risk and judgement roles that remain, while the people who fill them age out and grow scarcer. The pipeline that quietly produced senior talent is breaking in the middle.
This paper argues that the answer is not to automate harder or recruit harder, both of which are failing, but to build capability deliberately, using a skills-first operating model applied across three distinct parts of the workforce. And it argues that in this sector, uniquely, the system that builds that capability is the same system that proves it to the regulator.
PART ONE · THE PROBLEM
Unlike retail or manufacturing, this is knowledge work: credentialed, desk-based, and conducted inside one of the most heavily regulated labour markets there is. The workforce is smaller in headcount but far higher in value per head, and that changes the nature of the problem entirely. Losing a senior adviser with a client book, or a financial-crime specialist who knows your systems, is not a vacancy to backfill. It is a capability hole that can take a year and a fortune to close, if it can be closed at all. Two forces are widening those holes faster than the sector can fill them.
AI is reshaping cognitive routine work faster than physical automation ever reshaped the factory floor, and finance sits squarely in its path: loan and mortgage processing, KYC and AML operations, first-line analysis, customer service, paraplanning and back-office administration. International evidence already shows the mix shifting, with demand rising sharply for specialist, data and regulatory skills while clerical, administrative and entry-level vacancies fall, a pattern most pronounced at graduate level. Wall Street banks have signalled the elimination of large numbers of back-office and entry-level processing roles over the next few years as AI absorbs the work.
Here is the part most cost-focused automation programmes miss. Those entry roles were not just cost. They were the training ground. They were where a graduate learned how a loan actually gets assessed, how a client's needs surface, how the institution really works, before progressing into advice, lending judgement or risk. Remove the bottom rungs and you do not just save salary. You sever the path by which juniors became seniors. As one analysis put it, banks that automate without redesigning roles risk hollowing out institutional judgement, and the digital talent shortage is less a pipeline problem than an identity problem.
While AI thins the bottom, conduct and competency regulation has lifted the bar at the top, and made capability a legal obligation rather than a nicety. In Australia, the Hayne Royal Commission and the professional standards that followed helped cut the number of financial advisers from roughly twenty-eight thousand in early 2019 to around fifteen thousand by 2025, almost a halving, even as an ageing population's need for advice rises. In New Zealand, the Financial Advice Provider regime now requires every adviser to meet and continuously maintain competence, knowledge and skill under an enforceable Code, and the Conduct of Financial Institutions regime extends fair-conduct and capability obligations across banks and other institutions, with consumer lenders to be licensed from mid-2026.
The combined effect is a vice. The roles that produced tomorrow's advisers and specialists are being automated away, while the bar to perform and to legally hold the senior roles keeps rising, and the incumbents age out. Demand for the scarce specialisms, risk, compliance, financial crime, cyber and data, keeps climbing as each new regulation lands, from operational-resilience standards to expanded anti-money-laundering obligations. Australia now ranks among the world's most talent-short markets for exactly these banking and financial-services skills.
The squeeze lands differently on three distinct parts of the workforce. Solving it means treating them as three problems with one underlying cause, not one undifferentiated talent shortage.
Operations, contact centres, lending and KYC/AML processing, paraplanning. Most exposed to AI.
This is where AI bites first and hardest. Volumes of routine cognitive work, application processing, checks, first-line queries, are being absorbed by automation, and the headcount is falling and reshaping. The reflexive response is redundancy. The strategic response recognises two things: that the people in these roles hold valuable institutional knowledge and customer understanding, and that this tier was the entry point to everything above it. The question is no longer how many of these roles you need. It is which of these people can be reskilled into the work that remains, validating AI outputs, handling exceptions, exercising judgement, and into the advice and specialist roles the sector is desperate for.
Financial advisers, mortgage advisers and brokers, private bankers, relationship managers. Most exposed to regulation, trust and the ageing pipeline.
This tier is defined by conduct and competence regulation and by a demographic problem. The adviser workforce has shrunk dramatically and aged, just as intergenerational wealth transfer and a wave of retirements push demand for advice up. Every adviser must now prove and continuously refresh their competence to a regulator, and the firm is accountable for it. Yet most firms cannot answer basic questions with confidence: who is licensed and current for what, whose CPD is up to date, who is ready to step up as a senior adviser retires, and where the succession holes are. Trust, the entire basis of the relationship, now rests on demonstrable capability the firm often cannot actually see.
Risk, compliance, financial crime and AML, cyber, data, quant, credit and treasury. The war for talent.
These roles are in a global bidding war that regulation keeps intensifying, as each new obligation, operational resilience, anti-money-laundering expansion, AI risk, creates fresh demand. They cannot be filled fast enough, contractor premiums are steep, and migration only goes so far when every developed market is chasing the same people. The instinct is to keep buying them on the open market. The evidence that most firms already outsource these functions is proof that strategy has hit its limit. The durable supply is to build these specialists from the analyst and processing ranks below, which is precisely the pipeline that force one is eroding.
| Dimension | Processing & service | Advisory & relationship | Scarce specialists |
|---|---|---|---|
| The dominant force | AI automating routine cognitive work | Conduct & competency regulation, plus an ageing workforce | A global war for scarce, regulated expertise |
| Where it breaks | The entry roles that trained everyone are disappearing | Advisers retire faster than they are replaced; competence must be proven continuously | Roles cannot be filled fast enough; contractors cost a fortune |
| The reflex that fails | Automate and make redundant | Recruit experienced advisers from a shrinking pool | Keep buying talent on the open market |
| What's missing | A reskilling path as roles are reshaped | A visible competence-and-CPD record and a succession plan | An internal pipeline instead of a permanent hiring war |
Retention dynamics in this sector are not about hourly churn. People leave for progression, for scope, and increasingly for flexibility and wellbeing as much as pay, and the most damaging losses are the scarce and the senior: the specialist a competitor poaches, the adviser who takes a client book out the door. In a market this short of regulated expertise, keeping your scarce people is not an HR metric. It is risk management.
The single root, across all three tiers: most banks and advice firms cannot see the capability they already hold. Competence is scattered across CPD logs, training records, licensing registers and managers' memories, none of them joined up. So the firm cannot answer the questions that now matter most: who could be reskilled as AI reshapes their role, who is ready to advise or to move into risk, whose competence is current and provable, and who will replace the specialist or adviser about to leave. You cannot reskill, redeploy, build a pipeline, or prove competence you cannot see. And in this sector, proving it is not optional.
THE TURN
The three problems share one cause: the sector has relied for decades on capability being absorbed by osmosis, juniors learning the trade by doing the routine work, the best rising over time into advice and specialism. AI is dissolving that mechanism, regulation is raising the bar it used to feed, and demographics are draining the top. Capability can no longer be left to accumulate on its own. It has to be built on purpose.
That reframes the whole challenge. Automating around people, or recruiting your way out of a shortage the whole market shares, both treat capability as something you acquire elsewhere. A skills-first model treats it as something you grow, deliberately, from the workforce you already employ. In banking, wealth and advice, that approach has a second payoff found in no other sector: the live record of skills and competence you build to develop people is the same record you need to satisfy the regulator. The compliance burden and the capability engine become one system. That is what TalentJam is built to do.
PART TWO · THE SOLUTION
TalentJam is a skills intelligence platform built on a continuous loop. Skills feed Performance, Performance feeds Growth, and Engagement runs through all of it. The four disciplines apply across all three workforce tiers; what changes is which carries the most weight. In this sector the loop does something it does nowhere else: it turns the regulatory obligation to evidence competence into the very engine that builds it.
TalentJam builds a living skills and competence profile for every person, mapped to the roles you run today and the roles AI is reshaping them toward. In financial services that profile does double duty: it is both a development tool and a source of regulatory truth, holding competence, CPD currency and licensing readiness in one live system instead of scattered logs. It is the direct answer to the finding that firms are moving to skills-based decisions yet cannot reliably assess the skills they hold. In processing it reveals who can be reskilled rather than made redundant; in advice it shows who is current, licensed and ready; in the specialisms it pinpoints the single points of failure you cannot afford to lose.
Skills & competence profiles / CPD & licensing currency / Role & AI-shift mapping / Succession visibility
This sector already lives by evidenced competence: licensing, fit-and-proper, conduct obligations, CPD. TalentJam makes that native, with verifiable competence sign-off and light, regular, skills-anchored check-ins that sit alongside the conduct framework rather than bolted on once a year. Just as importantly, it equips managers to develop people, which matters more than ever now that the on-the-job training ground is eroding. When juniors can no longer learn by osmosis from routine work, deliberate development by a capable manager becomes the pipeline.
Competence sign-off / CPD tracking / Continuous feedback / Manager enablement
When the people who matter most are the hardest to replace, retention is everything, and today's professionals weigh flexibility, scope and wellbeing alongside pay. TalentJam's engagement capability provides low-friction listening and structured recognition that flags a disengaging specialist or a senior adviser before a competitor does, while there is still time to act. Losing one financial-crime lead or one adviser with a client relationship is not a soft cost. It is a capability and revenue hole.
Pulse listening / Recognition / Early attrition signals / Retention of scarce roles
This is the pillar that answers the squeeze. The advisers and specialists the market cannot supply have to be grown, and grown on purpose. TalentJam maps real progression, from processing or analyst roles into advice, with the competence and CPD steps the regime requires built in, or from analyst into risk, compliance, financial crime and data, and ties each step to the specific capability needed to climb it. That rebuilds the pipeline AI is eroding, converts capable people from reshaped processing roles into the work the sector is short of, and gives a sector competing for talent a reason for ambitious people to stay. It is the practical fix for the gap between the firms that plan to reskill and the far fewer that actually do.
Career pathways / Reskilling into advice & risk / Adviser succession / Specialist pipelines
Most firms already own fragments of this: a CPD register, an LMS, a licensing spreadsheet, a learning portal nobody finishes. They sit in silos, and capability falls through the gaps between them. The loop is the point. Skills and competence data make performance and regulatory evidence concrete. Performance reveals who is ready to grow into advice or specialism. Growth rebuilds the pipeline AI is eroding and regulation demands. Engagement keeps those people long enough for their competence to deepen, which feeds the next turn. Each pillar makes the others work harder, and the compounding lets a firm grow the capability it can no longer buy, while proving it to the regulator from the same system.
THE STRATEGIC PRIZE
This is the point that matters most to a large bank or institution, and it is the slowest to land and the largest in value. The scarce risk, compliance, financial-crime, cyber and data capability that regulation keeps demanding cannot be bought reliably. The shortage is structural, the competition is global, and the steep contractor premiums most firms already pay are proof that the open-market strategy has reached its limit. The alternative is to treat the analyst and processing workforce, including the people whose routine roles AI is reshaping, as the deliberate pipeline for the scarce roles above, and to manage that pipeline as carefully as any other strategic asset.
A firm that can see latent capability in its people, identify who can be developed toward financial crime, risk or data, and move them along a mapped path that builds and evidences the required competence, owns a supply of scarce, regulated expertise that its competitors are still trying to recruit. That is a slower build than a contractor and a longer conversation than a single hire. It is also the one that compounds, satisfies the regulator, and cannot simply be outbid.
IN PRACTICE
Consider a mid-market wealth and advice business, several offices, an ageing group of senior advisers, a paraplanning and processing team whose work AI is rapidly reshaping, and a constant, losing battle to recruit compliance and risk capability. Nobody can say with confidence where the firm's real capability sits, or whether it could prove competence at short notice. Here is how the loop changes the trajectory.
From osmosis to design
→ Quarter one. Everyone gets a skills and competence profile, and the firm sees capability on one screen: who is licensed and current for what, whose CPD is up to date, where the single points of failure are, and that three of its most senior advisers will retire within three years with no identified successor.
→ Quarter two. As AI absorbs routine paraplanning, capable people from those roles are identified and placed on mapped paths toward advice, with the competence and CPD steps the regime requires built in, instead of being made redundant. Competence evidence now lives in one system the regulatory-facing team actually trusts.
→ Quarter three. An engagement signal keeps a scarce financial-crime specialist a competitor was courting. Two analysts are placed on a mapped path toward risk and compliance, rebuilding from within the pipeline the firm could never recruit fast enough.
→ Year two. The firm meets its conduct and competence obligations from the same system it uses to plan capability. The adviser-succession gap is closing with internally grown advisers, and it is building the specialists it used to rent. Reskilling has moved from an intention to a pipeline.
The same loop turned AI from a redundancy exercise into a redeployment engine, closed an adviser-succession gap, and started building scarce specialists internally, all while making competence provable on demand. That is the point of solving the squeeze with one model rather than three disconnected fixes.
THE TIMING
The forces are not slowing. AI is reshaping cognitive work fastest of all, and the gap between firms that intend to reskill and those that actually build the pathways is precisely where talent strategies are failing. Regulation is not retreating: conduct regimes are bedding in, anti-money-laundering obligations are expanding, operational-resilience standards are landing, and consumer-lender licensing arrives in New Zealand in mid-2026. The adviser shortage collides with a retirement and wealth-transfer wave. And firms across the market are already shifting toward skills-based decisions, while admitting they cannot reliably assess the skills they hold. That assessment gap is the one skills intelligence closes.
The institutions that pull ahead over the next decade will not be the ones that automate fastest or bid highest for scarce talent. They will be the ones that can see, grow, keep and prove the capability the sector now demands, building advisers and specialists internally while the rest of the market fights over a pool that is too small and getting smaller. That is an advantage a competitor cannot buy back.
See your capability. Build it. Prove it.
TalentJam gives banking, wealth and advice firms a live picture of the capability inside their business, the loop to grow and keep it, and the competence record to satisfy the regulator, all from one system. To see what it looks like for your firm, visit www.talentjam.io to book a walkthrough.
SOURCES & NOTES
ASIC Financial Adviser Register and industry analysis (Rainmaker, WealthData, CPA Australia, 2025): decline in Australian financial adviser numbers since 2019 and demographic mismatch. Financial Markets Authority (NZ): Financial Advice Provider regime, Code of Professional Conduct competence and CPD requirements, Conduct of Financial Institutions regime, and consumer-lender licensing from 1 July 2026. Wolters Kluwer (2026): reskilling intent-versus-action gap, talent scarcity as the leading barrier to scaling AI, projected back-office and entry-level role reductions, and the institutional-judgement risk of automating without redesign. Morgan McKinley London Employment Monitor and 2026 Australia Salary Guide: shift in finance vacancies toward specialist, data and regulatory skills and away from clerical and entry-level roles, and resilient risk and compliance demand. EY and McKinsey: financial-services AI adoption and cost focus. Brookings (2025): generative AI and the cognitive displacement of middle-office work. Aon Global Risk Management Survey (2025): cyber, regulatory change and talent shortage as leading Australian business risks. Industry talent-market analyses (2025 to 2026): risk, compliance, financial-crime, cyber and data shortages, and Australia's standing among the most talent-short markets. Several figures are international or all-sector indicators applied to banking, wealth and advice where local data is unavailable, and are described as such. Figures cited as approximate or as ranges reflect variation across published studies and methods.