---

In the previous four blogs in this series, I explored how leading financial services firms are scaling AI for value, driving rapid change, using agentic architectures to reinvent work and businesses, and reshaping hybrid workforces. In this final blog, I step back to the bigger picture: how leadership, culture, and the operating model shift when AI is scaled across the enterprise. I also highlight several of the questions boards are now asking.

AI and leadership

Leadership is a decisive factor in AI outcomes. Our research found that CEO-sponsored gen AI-enabled enterprise reinvention delivers 2.5 times more value than efforts initiated further down the organization. This signals the importance of AI to the strategy and future of the company.

The best CEOs we work with treat AI as part of the business strategy and the broader investment and transformation portfolio. They connect AI to purpose, vision, and performance. They also build a coherent narrative that helps the organization understand what is changing, why it matters, and how to engage. (My colleagues in our Storytellers team support many clients on exactly this.)

Roles and accountability are changing

Most CEOs appoint someone “on point” for AI, often the CIO or CTO. In some firms, the Chief Data Officer has been elevated, or a Chief AI Officer has been added to the C-suite or extended leadership team. These roles bring expertise and accountability, but AI cannot sit with one function. It is the responsibility of the entire C-suite. Each leader must own how AI is used in their business, how risks are managed, and how outcomes are delivered.

The board’s role is becoming more material

The CEO and key leaders need alignment and support from the board. Board relationships and tech experience vary significantly. Our research shows that tech expertise among board members has grown from 6% in 2015 to 16% today. About 20% of banks now have more than a third of board members with technology experience and are more likely to be successful. Yet 17% of bank boards still have no members with tech experience, which makes support and oversight far harder.

In my experience, board sessions are vital. They create space to align on strategy, investment, and risk appetite. They also often produce some of the strongest advocacy for a human-led approach.

 AI and leadership development

AI changes how leaders lead. It shifts the pace of decision-making, the nature of work, and the shape of collaboration. Leaders and leadership teams need new knowledge, skills, and behaviors. The Turing Institute’s AI Skills for Business Competency Framework acknowledges leaders as a key persona.

Fluency, not specialism

Most leaders do not need to become AI specialists. They do need business fluency: an understanding of capabilities and limits, the major AI types, and how AI fits into data, technology, and business architecture. That fluency helps leaders spot opportunities, define problems well, and sponsor initiatives intelligently. It also grows fastest when leaders use AI tools themselves.

Investing and scaling with a reinvention mindset

Leaders must improve how they evaluate and scale AI investments. That includes building a balanced portfolio, prioritizing resources, and funding the change beyond technology. Scaling AI requires investment in process, work design, skills, adoption and governance, not just models and infrastructure.

Risk, resilience and trust at scale

Leaders also need to deepen their risk capability. As discussed in blog 3, they must set risk appetite with the board, assign accountability for AI products and use responsible AI frameworks, tools, and specialist support. This expands leadership responsibilities across legal and regulatory compliance, risk ownership within the three lines of defense and control behaviors in areas such as anti-money laundering and fraud prevention.

FS leaders must also engage proactively with central banks, regulators, and policymakers. We will only build trust at scale through evidence and embedded practice, particularly in high-stakes decisions such as credit, claims and trading. At the same time, leaders must strengthen resilience and security in a world of AI-accelerated threats, embedding AI into cyber defense, operational risk and stress testing.

Change leadership becomes a core capability

This is fast-paced, evolving change across colleagues, customers, markets, and society. Leaders must sharpen their change leadership. They must be advocates who bring the vision to life, build trust and activate adoption. They must collaborate across business and technical teams and across the ecosystem. They must lead with authenticity and inclusion, make hard decisions, and address concerns directly.

AI is also changing constantly. Curiosity and continuous learning are now leadership essentials. Over the last two years I have supported leaders from board and C-suite to frontline managers and their appetite to learn continues to impress me. People arrive with different starting points, but the leaders who progress fastest ask questions, use the tools and translate awareness into practical application.

If you would like to explore how we support leader capability building and real-world AI application, please reach out.

Moving from initial leadership education to the application of AI

Many leadership teams have already had an AI “101” briefing, or a “201” business school visit. That creates baseline literacy, and it matters. We now also have a full AI Academy and TQ modules within our LearnVantage capability to support this ongoing learning.

For most organizations, the leadership challenge has now moved onto applying AI to real business opportunities and problems and leading the change at scale across functions and teams.

Preparing and aligning middle managers and team leaders

AI changes management work. Coordinating, supervising, and controlling work looks different when agentic architectures become part of day-to-day operations. AI becomes more than a tool.

AI is not human and not fully autonomous, but it can sense, generate, remember, plan, and act. In a regulated industry, the ability to oversee performance, conduct, and decision-making across both human workers and AI agents will become increasingly important. Managers may need to supervise orchestration agents within agentic workflows.

Spans of control may expand sharply as “teams” include both humans and agents, moving from 10 to 100. To manage this, leaders will need dashboards and AI-enabled management tools for governance and control. The goal is to spend less time on admin and more time on judgement.

This is why leadership learning must include managers and team leaders, not only executives. These leaders need the same space to build understanding, ask questions, and start using AI. They also carry the day-to-day responsibility of addressing concerns, supporting adoption, and resetting ways of working. Aligning leaders at all levels materially improves transformation outcomes (Accenture Transformation GPS, 2025).

Evolving the culture for an age of intelligence

AI does not change a bank or insurer’s purpose and values. It does change how we work, think, and decide. Digital shifted culture in financial services; AI will shift it further and faster.

The specific cultural themes vary by firm, but common areas include:

• • Psychological safety in teams, and trust in AI.
  • Curiosity and learning new ways of working.
  • Integrating responsible AI compliance with innovation and speed.
  • Radical transparency about where and how AI is used.
  • A reinvention mindset that challenges the status quo.
  • Breaking siloes and building collective intelligence.
  • Protecting human value and human relationships at work.
  • Ensuring agents uphold conduct and consumer standards (for example, Consumer Duty in the UK).
  • Strengthening resilience and addressing systemic AI risks.

What enables that culture change?

First, tone matters. Leaders set the climate for trust and psychological safety, including the safety to speak up when something is wrong. I have been encouraged by leaders who discuss risks openly and treat concerns as legitimate.

Second, teams need empowering opportunities to take the first steps. The organisation should set a coherent view of how work will be done and then give teams space to experiment, learn, and adopt AI well.

Third, leaders must challenge siloes and “the way we’ve always done it.” Financial services has often had a bias to the status quo. Reinvention requires deliberate disruption.

Finally, culture follows systems. If we want new behaviours, we must change the operating mechanisms around them. Delivering with AI should be low friction.

Regulatory tone matters as well. A more innovation‑positive stance, such as the UK FCA’s direction, shapes firm policy, compliance expectations and audit priorities.

Over the last five years, we have lasting shift in the importance of human needs at work: wellbeing, purpose, growth, and inclusion. As we apply AI, we must strengthen these elements, not erode them. Financial services is at its best when it delivers meaningful work, strong relationships, trust and belonging.

Leading organizations reinforce these human elements as they scale AI. They use AI to personalize career development and learning pathways, helping employees grow through meaningful work aligned to strengths and goals. At Accenture, leaders ask a bold question: “Are our people better off for working here?” They measure whether employees are building marketable skills, maintaining wellbeing, and feeling purpose and belonging. According to our Talent Reinventors survey, when employees feel “better off,” they are 1.7X more adaptable to change, 1.9X more likely to trust the organization and 35% more likely to drive innovation.

Adapting the operating model

An operating model is the set of capabilities and how they work together to serve customers. As AI and agentic architectures change workflows at the core of banks and insurers, operating models must adapt. Our recent thought leadership focused on the operating model changes required for AI. This highlights the need for more adaptive structures and readiness for continual change (see blog 3).

This shift challenges traditional organisational boundaries. Value streams cut through siloes, and financial services is full of them. We also need to move beyond one-off use cases and AI as standalone models. Agentic AI is becoming a reusable capability integrated with people, process, data, and technology.

Many clients are now asking a bigger question: what does an “AI-first bank” or “AI-first insurer” look like? Several themes sit inside that question. Let’s explore a few of the big questions within this.

Changing customer experiences and value dynamics

Agentic AI can make banking and insurance more proactive, conversational, and relevant. It can turn one-off digital transactions into continuous interactions that remember preferences and needs. It can also create space for firms to be more human when it matters most: in moments of emotion (bereavement, fraud), relationship (private banking), and advice (financial planning).

In the near term, AI interaction may feel daunting for elderly and vulnerable customers. Choice and inclusive design will matter.

Customer choice will also expand. Some customers will use personal agents to direct deposits and payments across providers via open banking. Customer journeys may increasingly be brokered between bots as much as between humans. That will reshape distribution, pricing, and loyalty.

Firms will take different competitive positions. Some will try to own the agent-customer interaction. Others will compete on trust through proprietary, highly compliant agents. Others will build ecosystems and integrate widely.

These choices will reshape economics, work, and workforce across each institution.

Real world example — changing customer interaction with AI

At a large Asian bank, we redeveloped the mobile app using AI in the creative process (including synthetic customer testing) and embedded AI for personalization and automated digital content and messaging. Across millions of touchpoints, the gen AI redesign increased customer satisfaction, drove 50% higher usage, reduced marketing costs by 40%, increased digital sales by 65% and lifted cross-sell to 30%. These are not customer-owned agents yet, but it signals how quickly AI can improve experience and performance.

Growth models leveraging AI network effects

AI will gradually reshape competitive logic. As firms embed agentic architectures across the core value chain, they lay the foundations for AI-first growth models. Over time, these models may create AI network effects, where learning loops improve performance faster than traditional scale advantages.

Fraud and transaction monitoring illustrate the direction. As agents learn from patterns and from each other, they can identify new features and signals. Agentic architectures do not only automate. They can create learning systems that improve continuously and feed collective intelligence. We are early, but the operating model implications are significant.

Composable, resilient and sustainable operating models

Agentic AI points to three operating model requirements:

1. Composability: Financial services capabilities are deeply interconnected and often hard wired into legacy platforms. Composability means designing modular AI components and agents that are reusable across products and geographies. That enables faster reinvention and reduces technical debt.
2. Resilience: Firms must build human-and-agent operating teams that can absorb shocks, maintain integrity and recover quickly. In regulated sectors, continuity and trust are non-negotiable. This includes the ability to run essential services if an agentic layer fails.
3. Sustainability: As AI compute workloads grow, banks and insurers must manage energy, carbon, and water usage, model efficiency, and greener cloud architectures. These choices will influence net-zero goals and regulator, customer and investor sentiment. There is also opportunity to fund the sustainable energy and water solutions data centres require.

Alignment with wider operating model shifts

Agentic AI arrives while many banks and insurers are already simplifying governance, adopting product operating models for continual change and evolving global capability centers (GCCs) into intelligent operations.

Product owners will need to evolve AI capabilities continuously and bring AI skills into delivery squads alongside engineers, customers, proposition and change professionals. GCCs are increasingly combining scaled operations with deep technology, data and AI capability to deliver intelligent operations for enterprise reinvention. Organizations without established capability centers may instead choose to onshore or transform operations in place.

Real world example — AI inside the GCC

With a large complex-lines insurer, we deployed more than 100 RPA, AI, and now agentic AI solutions across value streams and processes within the GCC. This improved service, insight and speed, and delivered a 35% cost reduction. Critical factors included consolidated work visibility, a front‑to‑back ambition, strong talent and sustained investment in both core platforms and AI.

Data and AI within the operating model

Historically, many firms treated data as a technology problem, funded primarily for remediation and regulatory compliance. That view is reversing.

Banks and insurers are increasingly treating data as a strategic business asset. Curating it has value in its own right, and it is also essential for training AI accurately and scaling it responsibly. As a result, we see rising investment in data and AI capabilities and greater emphasis on business and customer domain knowledge for data professionals.

Real world example: NatWest and AWS

We announced a five-year collaboration with AWS and NatWest Group to accelerate modernization of digital, data, analytics, and AI capabilities for its 20 million customers. The collaboration includes investment in colleague skills and a data driven culture. Paul Thwaite, NatWest CEO, said: “Our industry — and the expectations of our customers — are changing rapidly and we are building our capabilities in order to understand and serve their needs better and faster than ever before. Equipped with high quality data, we can continue to quietly revolutionize how we serve our customers through the use of AI and other technologies in order to provide more personalized products and services as a trusted partner in the moments that matter most.”

In closing: A constrained or an unconstrained future

This is a leadership moment. C-suites and boards face a clear choice between a constrained and an unconstrained view of what AI can mean for their organisation.

The constrained view aims to “get 10% out.” It focuses on narrow task automation, removing lower-value work, and optimising the workforce. It may also default to “not yet” or “fast follower.” Given market pressure, uncertainty, and investment fatigue, this response is understandable. But it is ultimately limiting.

The unconstrained vision is different. It is not about incremental savings, but about unlocking 10x performance and growth. In this model, growth is no longer limited by how many people a bank can hire. One person can lead and orchestrate a team of AI coworkers to deliver exponentially greater impact.

This is a deliberate choice to reinvent the business by combining human and AI capability. It assumes continual change, aligned leadership and deep trust, with people leading the transformation. Success depends on empowering employees to reimagine workflows and co‑design intuitive human–AI interactions that elevate work rather than replace it. The central question shifts from “Where can AI cut costs?” to “Where can intelligence transform outcomes?”

The next generation of leadership will not be defined by how much control we retain, but by how responsibly we accelerate capability, both human and artificial. The banks and insurers that thrive will be led by people who combine purpose, curiosity and courage and who elevate intelligence for the benefit of customers, colleagues, shareholders and society.

 Key Takeaways

1. 1. Leadership: How are you helping leaders understand AI, apply it to business opportunities, role model its use, and lead their people through change?
   2. Culture: Culture shapes AI use, and AI use shapes culture. Responsible AI changes risk and control ownership. Are you actively considering AI’s cultural impact?
   3. Operating model: Is your operating model designed for continual change? How will you adapt the three lines of defense for AI risks? Do you treat data and AI as commodities, or as distinctive capabilities?
   4. Mindset: Agentic AI is a leadership choice. Does your leadership team have a reinvention mindset, or a constrained one?

Conclusion

Over the last five blogs we’ve covered a lot of ground:

1. 1. The competitive advantage of scaling AI: How AI and agentic architectures are reshaping financial services.
   2. Taking a human-centered approach to AI: How AI and agentic AI in particular help us reinvent work and reshape the workforce in financial services.
   3. Leading the change for value: Strategies for driving successful AI transformations, as true business change that is scaled, responsible, distributed, rapidly evolving and human-led.
   4. Workforce change and the role of HR: How agentic AI is reshaping the workforce and creating a critical role for the HR function.
   5. Leadership, culture and operating model change: How AI is starting to change how we lead, ways of working and become a fundamental capability across our industry.

AI continues to be one of the most disruptive and exciting changes within financial services and the wider world of work. There is a gap emerging between the leaders who have scaled AI for value and taken a human-led approach.

I hope the practical examples and my reflections have been helpful to you in bringing to life some of the approaches that can make the difference. I’d love to chat further with you about your opportunities and challenges — please reach out to me directly on LinkedIn and read our Banking Top Trends for 2026.