Financial services have been at the forefront of adopting new technology to reimagine how work gets done. From artificial intelligence (AI) to cloud computing, firms are changing business models, processes and roles. And what those changes mean for their workforce is now firmly on the C-suite agenda.

Recent research from Fitch Learning points to the scale of the shift:

• 94% of financial institutions expect routine tasks to be automated.
• 76% expect roles to be redefined through collaboration.
• 71% anticipate existing roles being partially reshaped.
• 31% expect entirely new roles to emerge.

For learning and development (L&D) teams, these changes create a complicated challenge. There is no single learning strategy that can address every type of role change. Some employees will need to learn how to work effectively with AI. Others will need to prepare for new roles as routine tasks are automated. And increasingly, organizations will need to develop the human capabilities that become more important as AI takes on more analytical and routine work, including critical thinking, communication and particularly judgment.

Three Patterns of Role Change — and Three Different Learning Needs

AI is changing jobs in different ways, and each type of change creates different learning needs.

• Augmentation: In augmented roles, employees work alongside AI and need to combine technology skills with stronger human judgment. For example, a research analyst might use AI to build a business case for a build-versus-buy decision, evaluate the trade-offs and then practice presenting the recommendation to decision-makers. The technology can accelerate the analysis, but the employee still needs to determine whether the recommendation makes sense, what assumptions need to be challenged and how to communicate the decision.
• Automation: In roles where AI significantly reduces or eliminates routine tasks, employees may need to reskill into adjacent capabilities. An operations analyst, for example, may spend less time reconciling data as those tasks become automated and more time reviewing exceptions, investigating anomalies or making decisions that require human judgment. L&D can support that transition by identifying the capabilities employees will need in the redesigned role and creating a clear path to proficiency.
• Newly created roles: AI is also creating roles that combine technical, business and risk-related skills that may not exist in traditional competency frameworks. An AI oversight analyst, for example, may be responsible for reviewing AI-generated content for accuracy, consistency and fairness before it moves to the next stage of a process. Preparing people for these roles requires more than adding AI training to an existing curriculum; organizations may need to rethink the underlying skills and competencies required.

These changes also affect groups that can be overlooked in AI learning strategies:

• Senior leaders need the ability to interpret AI developments, make strategic decisions and lead through change, yet they are not always the primary audience for AI learning programs.
• Control functions, including risk, compliance and audit, need enough technical understanding to effectively challenge AI-driven decisions and processes.
• Support functions, including HR, operations, legal and finance, need to understand how AI changes their own workflows and the processes they support. For example, HR may need to rethink how it conducts talent reviews or maintains skills and capability frameworks as AI becomes part of everyday work.

The common thread is that AI adoption is not simply a technology challenge. It is a workforce capability challenge.

Judgment and Responsible AI

One of the biggest risks of AI adoption is overreliance. People can be inclined to accept AI-generated analysis too quickly, sometimes giving machine-generated recommendations less scrutiny than they would give a colleague’s work. In financial services, that can have serious consequences. An unchecked recommendation can affect regulatory compliance, credit decisions, market activity or operational risk.

That makes judgment an increasingly important capability to develop — not simply a trait employees are expected to acquire through experience.

One challenge is that AI may remove some of the experiences through which employees traditionally developed judgment. Early-career professionals have historically built expertise by doing the groundwork: analyzing information, working through routine decisions, making mistakes, receiving feedback and gradually learning to recognize patterns.

As more of that work becomes automated, employees may have fewer opportunities to build that experience. As a participant noted in the research: “How do you teach wisdom? How do you teach people in their third year the things you’ve picked up after 20 years?”

The question is particularly relevant as AI takes over more of the analytical work traditionally assigned to junior employees. Those employees may still be expected to evaluate AI-generated recommendations, identify errors and know when a result requires further investigation even though they have had fewer opportunities to develop that judgment themselves.

This is where practice becomes important. Rather than treating judgment as something employees simply acquire through experience, organizations can create opportunities to practice it in realistic situations. Simulations, for example, can put employees in scenarios where the information is incomplete, competing priorities exist and there is no obviously correct answer. Learners can then receive feedback on not only what decision they made, but how they arrived at it.

The goal isn’t to teach employees to distrust AI. It’s to help them recognize when AI output is useful, when it needs to be questioned and when human expertise needs to take over. That distinction is becoming increasingly important as AI becomes embedded in everyday financial services work.

Judgement and the Future of Learning

If judgment is the capability firms need to build, then how learning gets delivered has to change, too. Five shifts are shaping the next generation of workforce development in financial services, and each one changes how judgment is developed and measured.

1. Learning in the flow of work

Pulling people out of their jobs for occasional training no longer reflects the pace of change. Development of judgement needs to happen in the flow of work, at the moment it is needed. Learning on the job with an AI coach that can role-play a difficult customer discussion, a performance meeting with one’s manager or influencing a group of colleagues, becomes a game-changer in the workplace. AI-enabled simulations are becoming one of the most effective ways to do that.

2. Personalization at scale

Organizations need learning that feels relevant to the individual while remaining consistent across the enterprise. That means adapting to role, level, pace and capability gaps, while still defining a common standard that can scale. In practice that looks different by role — a relationship manager might use an AI-enabled simulation to rehearse a difficult client conversation where AI has flagged a portfolio risk, while a credit analyst may use an AI-enabled simulation to work through the exception cases the models can’t resolve on their own. Same judgment muscle, different context.

3. Language and cultural fit

Global firms cannot build capability consistently if learning only works in one language or one context. Effective scale requires content and delivery that can adapt across geographies and local learning needs. A practical place for learning leaders to start is piloting content in two or three markets with an AI-enabled simulation, using local reviewers to pressure-test the scenarios and letting regional teams adapt the delivery rather than just translate it.

4. A steeper learning curve

Time to proficiency matters more than ever. As roles evolve in quarters rather than decades, speed to desk-readiness is becoming one of the clearest measures of learning impact. Staying ahead of that curve means reviewing role profiles quarterly instead of annually, building shorter modular content that can be refreshed quickly, practicing using an AI-enabled simulation in the flow of work and staying close to the business so skill shifts get spotted before they show up in performance data.

5. ROI measured in proficiency, not completion

Training completion rates say very little about whether someone can make a sound decision under pressure. What matters is whether they can demonstrate both proficiency in the subject matter, and exercise judgment in realistic scenarios. Simulation makes that measurable. Every decision a learner makes in an AI-enabled simulation gets captured — what they chose, what they overlooked, how long they took to act — which gives leaders a picture of readiness beyond a completion certificate.

A Call to Action

Financial services organizations need to invest in training their workforce to know when to trust AI and when human judgment must take over. That capability is built through practice in realistic AI-simulated environments, tailored to the individual, embedded in the flow of work, and delivered at the scale and consistency organizations require.

As the velocity of technologically driven change increases, firms investing in judgment now are preparing their workforce to adapt for the future.