info@meradia.com
.svg.png){kind=small}
Executive Summary
The financial services industry has always embraced technology to increase speed, scale, and efficiency. Generative and agentic AI represent the next most consequential step in that evolution. AI can automate repetitive work, accelerate analysis, improve decision support, and fundamentally reshape operating models. But there is a less obvious critical risk.
AI and automation could eliminate the important experiences through which people learn how our industry works.
The strategic question for financial services leaders and Chief Human Resources and People Officers: How do we capture the productivity benefits of AI without damaging the human-capital pipeline that produces tomorrow’s experts and leaders?
AI Is Changing the Bottom of the Career Ladder
Financial services careers have generally followed a predictable progression. Analysts became associates; associates became vice presidents; and some eventually became managing directors, portfolio managers, chief investment officers, and operating executives.
The work at the bottom of the ladder is not glamorous. Junior professionals reconciled transactions, collected and validated data, populated models, reviewed documents, prepared reports, investigated exceptions, and built presentations. These activities are operational tasks, but they were also apprenticeships.
Through them, people learned how markets behave, how data moves through an organization, why controls exist, how experienced professionals make decisions, and when something that appears technically correct simply does not make sense. AI is increasingly capable of performing many of these activities.
McKinsey’s 2026 research describes this challenge: research, documentation, data cleanup, preliminary analysis, and other foundational activities are increasingly being streamlined or absorbed by AI. Yet these are also the activities through which early-career employees historically developed expertise and judgment.
Stanford researchers examining U.S. payroll records found that employment among workers ages 22–25 in highly AI-exposed occupations is approximately 19% below where it would have been had employment tracked comparable workers in less AI-exposed occupations. Stanford does not find evidence of broad economy-wide AI displacement. Instead, the effects appear concentrated among younger workers and occupations where AI substitutes for tasks rather than complements employees.
We’ve Seen this Movie Before
Remember offshoring? During the offshoring wave of the 1990s and 2000s, institutions moved substantial operational activity to lower-cost locations. The economics were compelling, and many of those decisions were rational. But we eventually discovered that many of the activities being moved were more than processes. They were training grounds through which employees learned securities processing, investment operations, data management, accounting, controls, and the underlying mechanics of the business.
Strategic Risk Is Larger Than Headcount
From a near-term financial perspective, AI adoption is highly attractive. If AI allows an organization to perform the same work with fewer people, productivity increases and operational margins expand. But workforce optimization and workforce strategy are not the same thing.
PwC’s 2026 Financial Services Workforce AI Survey illustrates the issue. Almost 8 out of 10 financial services leaders expect their workforce to shrink by at least 20% over the next five years, yet only 42% report having conducted enterprise-wide modeling of how AI will change labor capacity. There is a fundamental difference between calculating how many people AI may allow an organization to remove and determining what workforce the organization will need afterward.
Three long-term risks require our collective attention.
Leadership pipeline. Today’s analysts and associates are tomorrow’s senior professionals. Materially reducing entry-level opportunities will create capability and succession gaps that may not become apparent for years. If analysts no longer prepare investment memos, where will future portfolio managers learn how investment decisions are structured and defended?
Institutional knowledge. Much of what makes a financial institution effective is tacit knowledge: historical context, judgment, relationships, exceptions, lessons from previous market cycles, and an understanding of why processes evolved as they did. Technology can help preserve that knowledge, but only if organizations deliberately capture, preserve, and transfer it. How will junior professionals gain the insight to know when a NAV is reasonable or how to question ‘does it make sense’?
Culture and professional judgment. Employees learn and contribute to their organizations through other people. They observe how experienced professionals respond when markets move, clients become concerned, models fail, data conflicts, or controls identify something unexpected. After years of AI-automated exception processing, who will have learned the pattern-recognition skills needed to identify a novel failure when AI encounters a scenario it has never seen before?
Removing interaction between generations of employees weakens the mechanisms through which judgment and culture are transmitted. We need new AI operating models that redesign how work gets done without dismantling how people learn, develop, and ultimately lead.
From Workforce Reduction to Workforce Reinvention
The strategic objective should not be preserving every task or every existing role. That would sacrifice much of AI’s potential. Instead, organizations should distinguish between work that can disappear and experiences that cannot. A repetitive task may no longer require a person to perform it. But if that task historically taught an employee how the business operates, organizations need another mechanism through which that learning occurs. We can no longer assume early career development will occur naturally through work. We need to design it intentionally.
Instead of eliminating junior roles because AI can perform their traditional work, financial institutions need to reinvent those roles around AI-enabled analysis, critical thinking, exception management, judgment, client interaction, problem-solving, and exposure to decision-making.
The question changes from asking: What work remains for junior employees after AI? … to asking: What experiences must junior employees have to become the experts and leaders we will need five, ten, and twenty years from now?
A Responsible AI Integration Agenda
At Meradia, an F2 Strategy company, we believe AI transformation should be treated simultaneously as a technology, operating-model, and human-capital transformation. Our Responsible AI Integration approach is organized around several interconnected objectives.
- Understand human-capital impact. AI initiatives should be evaluated not only for efficiency and financial return but also for their effects on roles, skills, career pathways, and organizational capabilities.
- Redesign work around human-AI collaboration. Entry-level and professional roles should evolve, so AI performs appropriate repetitive activities while people increasingly develop the analytical, interpersonal, creative, and judgment-based capabilities that create differentiated value. Equally important, AI-enabled roles must be designed to engage and inspire junior professionals, that gives them meaningful opportunities to learn, contribute, and build the capabilities that will define the next generation of leaders.
- Build future human + AI capabilities. Organizations need deliberate strategies for reskilling and redeploying employees as the mix of human and machine work shifts.
- Protect institutional knowledge. AI transformation should strengthen mechanisms for transferring expertise, organizational context, and judgment across generations of employees.
- Govern workforce transformation. AI governance should incorporate workforce, talent, risk, and succession considerations alongside technology, security, regulatory, and model-risk requirements.
- Measure long-term organizational health. Productivity and margin improvement remain essential, but leaders should also understand whether AI is strengthening or weakening the organization’s talent pipeline, institutional knowledge, and future leadership capacity.
The objective is to capture AI’s economic value while preserving the human capabilities required to sustain and expand that value.
AI is creating a similar scenario and with a similar strategic choice … but at much greater speed and scale. This time, the question is not where the work moves, it is whether a human performs it at all.
Our industry is exceptionally good at pricing risk. AI introduces another risk that deserves the same discipline: human-capital risk. The managing directors, chief investment officers, technology executives, heads of operations, data leaders, risk officers, and investment professionals who will run our institutions in 2040 are entering the workforce today.
Their careers will look different from ours. AI will allow them to learn faster, analyze more information, automate routine activities, and operate with capabilities previous generations never possessed. But they still need opportunities to develop judgment. They still need exposure to experienced professionals. They still need to understand how the institution actually works. And they still need a first rung.
The choice is therefore not AI or people, but whether we use AI primarily to remove work or use it to reinvent how the next generation learns to do higher-value work. The organizations that get this right can achieve something far more valuable than short-term cost reduction: an AI-enabled operating model that simultaneously increases productivity and accelerates the development of human expertise. Every AI business case should answer two questions. How much efficiency will be created? And how will the organization continue developing the expertise it will need a decade from now?
AI’s success should not be built by eliminating rungs, but instead by how much higher we enable the next generation to climb.
References
Rethinking talent development in the age of AI | McKinsey
Financial services AI workforce gap: PwC
Download Thought Leadership Article Process Design and Change, System Rationalization Artificial Intelligence and Machine Learning, Data and Digital Transformation Asset Managers, Asset Owner, Wealth Managers Brian Buzzelli