How AI in financial services is transforming operating models

Firms are rethinking how work gets done, elevating human judgment and redesigning processes as intelligent systems deliver speed and scale.

In brief

• AI is shifting financial services toward new operating models centered on human-AI collaboration.
• Success depends on workforce readiness, trust and strong data governance.
• Leaders must integrate AI strategically while upskilling employees for new roles.

EY Americas Financial Services AI and Data Leader Phil Andriyevsky recently sat down with Massachusetts Institute of Technology (MIT) Professors Alex Pentland and Hossein Rahnama to address the transformative impact of artificial intelligence (AI) on the financial services industry and discuss the essential collaboration between human insight and AI capabilities in creating a more adaptive workforce.

Financial services is entering its most consequential operating model shift since the transition from paper-based ledgers to electronic trading. AI is no longer a peripheral productivity tool; it is becoming the core engine driving speed, scale and margin expansion in an industry that moves trillions of dollars a day. Across AI in financial services — and more broadly AI in finance — the firms that will emerge as winners are those that treat their workforce as a portfolio, skillfully integrating human capital and digital labor with the same rigor applied to capital allocation. As compute, energy and model performance evolve into competitive moats, the valuation multiples over the next decade will hinge on one critical factor: the ability of institutions to cultivate a workforce that fully leverages AI while maintaining culture, governance and trust.

The shift from automation to AI-driven operating models

This transformation is not merely about adopting AI technologies; it requires a fundamental shift in how the workforce operates. The rise of AI in financial services represents a move from traditional automation, creating a cohesive system where human insight and AI capabilities work in tandem to generate value. As leaders in AI, supported by their financial services partners, lay the groundwork for an expansive AI compute capability, they anticipate a future where processing and energy capacity are pivotal in reshaping the global economy and society.

However, the heavily regulated nature of the financial services industry presents significant challenges to this promise of innovation. While AI tools have enhanced compliance and streamlined operations, they also introduce new risks and governance challenges, particularly in an industry where trust is paramount.

AI in the workforce: the real constraint on scale

Technology is ready; however, in most cases, the workforce is not.

From generative AI to agentic AI in financial services

The financial services industry has decades of experience with automation, ranging from risk models to fraud detection. However, the advent of generative AI (GenAI), which focuses on content creation and the generation of new ideas, and agentic AI, which emphasizes autonomous execution and decision-making, signifies a structural shift from merely enhancing existing processes to replacing and reallocating entire workflows. This distinction is crucial, as GenAI enables organizations to produce content and insights at scale, while agentic AI empowers systems to act independently, executing complex tasks without human intervention.

Why AI in banking is moving from pilots to production

A recent EY-Parthenon study from Ernst & Young LLP highlights this momentum, revealing that 95% of wealth management firms are deploying at least three GenAI use cases. In banking, 89% of respondents expect strong positive impacts from GenAI within the next two years, while 69% of insurers are already looking forward to agentic AI, particularly in risk assessment and mitigation. This shift reflects a broader pivot across financial services from back- and mid-office use cases to a focus on how AI can drive revenue and enhance customer satisfaction. These new investments signal growing confidence in AI, moving exploration and investments from low-risk, internal-facing applications to frontline client interactions.

Despite these enhancements, challenges remain that hinder value creation and return on investment. According to recent MIT research examining the GenAI divide, as many as 95% of pilot programs fail to make it into production.¹ Key reasons for this include:

• Tools that are too generic and not customized for specific needs
• Workflows that are not adapted to new paradigms
• Resistance to changing established processes

Consequently, while financial services firms are adopting AI tools at a high rate, they are not experiencing the transformative change they anticipated. At the heart of this issue are human and process factors, including resistance to changing workflows, the complexity of evolving new processes, and the need for effective change management and upskilling.

Embracing workforce transformation: key actions for financial services leaders

As financial services firms transition to an AI-powered operating model, leadership plays a critical role in guiding teams through this transformation. Leaders must not only embrace technological advancements but also foster collaboration between human employees and AI agents.

Key imperatives include the following:

1. Implement a strategic integration plan: Anchor AI adoption in a defined operating roadmap, with visible leadership sponsorship, clear communication and feedback loops tied to business outcomes. Effective change management is essential to normalize new ways of working and sustaining momentum. Upskilling must be continuous and role-specific — combining practical AI fluency with human capabilities, such as problem-solving, adaptability and judgment. As AI access becomes increasingly democratized, organizations have a responsibility to equip nontechnical workers with the skills and mindset required to collaborate effectively with intelligent systems. Ethics and responsible AI training should be embedded throughout to reinforce trust with employees, clients and regulators.
2. Redefine performance management: Shift performance metrics from individual achievements to a focus on team collaboration and innovation. Implement personalized training programs tailored to specific roles, empowering employees to leverage AI effectively. Redefine roles to emphasize strategic thinking and relationship management, preparing the workforce for future challenges. Develop long-term career management strategies that provide clear pathways for growth within the organization.
3. Develop tailored training programs: Move beyond generic training modules so that training is role-specific. Balance technical expertise in AI tools with essential soft skills, such as communication, problem-solving and emotional intelligence. A holistic training approach will cultivate a workforce capable of navigating the complexities of human-AI collaboration. According to the EY Agentic AI Workplace Survey, 89% of desk workers believe upskilling is crucial, yet 59% cite inadequate training as a barrier to AI adoption. Leaders must prioritize comprehensive training programs that address these needs, harnessing the strong desire for development among employees.
4. Redefine roles and skills: As AI automates routine tasks, employees must adapt to roles emphasizing strategic oversight and relationship management. The “humans-in-the-loop” model is essential in this context, as it allows for human oversight in critical decision-making processes, such as signing off on AI-generated credit risk assessments. This oversight not only meets regulatory requirements for explainability but also enhances trust in AI systems because human judgment is integral to the decision-making process. Encourage teams to leverage AI for data analysis, allowing human analysts to focus on interpreting results and making strategic decisions. This collaborative workforce model will enhance decision-making and foster innovation, enabling employees to engage in higher-value tasks that require human judgment and creativity.
5. Build trusted data infrastructure: Treat data as a board-level asset class, verifying that it is accurate, governed and reconciled across the organization. Establish a firmwide source of truth with standard taxonomies and aligned metadata. Implement governance structures built for real-time AI, incorporating continuous monitoring and automated quality controls to meet rising regulatory standards. Transparency and auditability are non-negotiable, and AI should not be deployed without full data lineage, documented logic and audit trails.

By prioritizing these strategies, leaders in financial services firms can cultivate a dynamic workforce that thrives in partnership with AI, driving growth and innovation in an increasingly complex landscape.