How to Successfully Structure the Human + AI Banking Workforce

By James Papadopoulos, head of Americas for Fitch Learning

Published on June 9th, 2026 in Leadership & Management

Simple Subscribe

Subscribe Now!

The architecture of modern banks is rapidly changing. Agentic AI is moving into the operational core of banks, automating back office processes, generating analyses, interacting with customers and even executing business decisions. These AI agents autonomously complete tasks, make recommendations and coordinate with other software agents.

A 2025 EY-Parthenon study found that 77% of leading retail and commercial banking firms have launched or soft-launched GenAI and 31% have begun implementing agentic AI. A McKinsey study from the same year found that banks that have done AI pilots have seen a 20-40% reduction in cost to serve from AI augmentation of frontline teams.

Reality check: While AI has the potential to transform the banking industry, it also raises questions about accountability, risk and team design. Banks that get it right will unlock scale, personalization and responsiveness without sacrificing trust. Those that get it wrong risk poor outcomes, regulatory scrutiny and damaged customer relationships.

Key takeaway: It all starts with building a workforce that can successfully integrate humans and AI agents. Equally important is understanding how AI will augment — not simply replace — individual humans, elevating what people can accomplish and reshaping the nature of every role in the bank.

Structuring Human + AI Teams and Accountability

When not all of your contributors are human, you must think differently about team structure and accountability.

Like any workforce, it is critical to understand the strengths and weaknesses of the players. AI excels in improving efficiency and doing high-volume or repeatable tasks such as data synthesis, scenario modeling, document drafting and building financial models. Where it falls short is in uniquely human capabilities like persuasion, negotiation, trust-building, ethical judgment and empathetic customer service.

AI cannot be accountable, and this is why an intelligent organization must be led by people. Humans should remain the arbiters of quality and be the final decision makers where outcomes matter. They are also solely responsible for legal and regulatory matters. That means banks need to design processes where AI augments repetitive work while humans supervise, validate and deliver the relational value that drives growth.

Best practices include:

Re-think your role descriptions: Creating human + AI roles can help define what humans should bring to the table – skills such as customer relationship management, negotiation, decision-making, etc. – and those best performed by AI. AI agents also need accountable human owners and an escalation path for issues.
Define decision thresholds: Identify up-front which decisions agents will be allowed to make autonomously (e.g., routing a support ticket) versus those that will require human sign off (e.g., credit exception, pricing overrides) and determine thresholds for situations that should trigger a compliance review or human approval. Embed ethical and compliance checks into decision loops.
Standardize human in the loop protocols: Specify when and how humans must intervene, what override authority they have, and how feedback is captured to improve agents.
Invest in upskilling: Working with AI agents requires a different kind of critical thinking, different skill sets and different coaching and training needs. Skills such as client service, judgement, relationship building, and empathy will become even more important and should be prioritized and measured. In addition to re-thinking role descriptions, organizations will need to consider new career path plans and performance metrics.

-- Article continued below --

How AI Augments the Human Workforce

The most immediate and practical impact of AI in banking is not replacing employees — it is making each employee significantly more capable. Think of AI as an always-on analyst, researcher, writer, and problem-solver working alongside every member of your team. A relationship manager who once spent hours preparing client briefings can now do it in minutes. A credit analyst can run scenario models in real time during a client call. A branch manager can receive AI-generated coaching insights after every customer interaction. The employee remains the decision-maker and the relationship owner; AI simply removes the friction.

Key insights: Three augmentation patterns are emerging across financial institutions:

First, AI as a thoughtful co-pilot, surfacing relevant data, summarizing complex documents, and drafting initial outputs that professionals then refine and own.
Second, AI as a real-time coach, providing employees with prompts, compliance alerts, and next-best-action recommendations during customer interactions.
Third, AI as a capacity multiplier, allowing teams to handle larger portfolios, more complex queries, and faster turnaround times without proportional headcount increases. Banks that deploy AI in these ways are not downsizing their teams — they are upgrading their capabilities.

A New Era in Risk Management

An AI-driven organization opens up new risks as well as new rewards. Risks can be managed but this requires planning ahead and making changes and updates to risk management practices and procedures. Some key steps include:

Establish an AI oversight function: This function should report to a senior executive (CRO, COO or chief transformation officer) that enforces governance, model inventory and auditability.
Define Service Level Objectives (SLOs) for agents: Accuracy thresholds, escalation rates, customer satisfaction metrics and time to remediate incidents.
Create incident playbooks: Prepare for model failure modes with clear roles, communication protocols and remediation steps to protect customers and reputations.
Control third party risk: Apply due diligence to vendor models, require transparency about training data and controls, and include contractual rights to audit and change.
Update liability and contractual frameworks: Ensure legal, compliance, and vendor teams are embedded in program design from day one.
Monitor for bias and hallucination: Run bias tests, red team models, and business outcome checks to detect drift and false positive/negative behaviors.
Communicate transparently around AI usage: Explain to customers how AI is used, what protections exist, and how they can escalate issues.

The Rise of the AI-Driven Intelligent Organization

The AI-driven Intelligent Organization in which humans and AI agents work interconnectedly has arrived. It is changing not only how companies operate but also entire organizational structures. For financial services and HR leaders, this creates an urgent need to redesign governance, talent, operating models, and learning pathways so that employees can work effectively alongside AI.

Banks that move quickly can build more adaptive, innovative organizations that operate at market speed. Those that do not risk regulatory gaps and issues, operational breakdowns, and talent attrition.

The industry is banking on AI and those that adapt earlier — and intelligently — will get ahead.

The New Skills Imperative: What the Intelligent Organization Demands

The rise of the AI-driven Intelligent Organization does not diminish the importance of human skills — it intensifies it. As AI absorbs more of the analytical, transactional, and administrative workload, the distinctly human capabilities become the primary source of competitive advantage. Banks that invest in these capabilities now will be better positioned to attract talent, retain customers, and build the kind of institutional trust that no algorithm can manufacture. The following skills are not soft in the sense of optional; they are the critical professional competencies for success in the AI era.

Critical thinking and AI oversight: Employees must be able to evaluate AI outputs with rigor — identifying errors, detecting bias, and recognizing when a model’s recommendation does not align with contextual reality. This is not about distrust of AI but about responsible stewardship. Banks need professionals who can interrogate an AI’s reasoning, not just accept its answers.
Emotional intelligence and empathy: As AI handles the routine, every human interaction carries greater weight. Customers who speak with a banker are increasingly doing so because the matter is complex, sensitive, or high-stakes. The ability to read emotional cues, respond with genuine empathy, de-escalate tension, and build psychological safety in a conversation is not something AI can replicate — and it will define the quality of customer relationships in the intelligent organization.
Influence, persuasion, and negotiation: AI can surface the best options; humans must still persuade stakeholders to act on them. Whether it is a relationship manager winning a competitive mandate, a credit officer advocating for a nuanced exception, or a team leader aligning colleagues around a new AI-enabled process, the ability to build consensus and move people is indispensable. These skills require practice, coaching, and structured development — they do not emerge by accident.
Ethical judgment and accountability: In an environment where AI can generate recommendations at speed and scale, human ethical judgment is the essential guardrail. Employees need to be equipped to ask the harder questions: Is this fair? Does this serve the customer’s long-term interest? Could this create unintended harm? This requires a combination of values clarity, domain expertise, and the confidence to escalate concerns even when outputs appear statistically sound.
Adaptive communication: The ability to translate complex, AI-generated information into clear, confident, and accessible language — whether for a customer, a regulator, a board, or a colleague — is increasingly valuable. With AI producing more of the analytical content, the human role shifts toward interpretation, storytelling, and connection. Professionals who can communicate with precision and clarity across audiences will stand out.
Continuous learning agility: AI tools, workflows, and expectations will continue to evolve at pace. The capacity to learn new systems, unlearn outdated habits, and adapt to change without losing confidence or effectiveness is itself a skill that must be measured and developed. Organizations that make learning a cultural norm — not a periodic event — will be most resilient to the next wave of disruption.

Building these capabilities requires intentional investment: redesigned learning programs, updated performance frameworks, coaching cultures, use of AI simulation platforms for training, and leadership that models curiosity and openness to change. The Intelligent Organization is not built on technology alone — it is built on people who are equipped, empowered, and continuously developed to thrive within it.