Interview with Edle Everaert, Partner and Global Head of AI, Projective Group
Q1. You’ve spent over 25 years advising financial institutions on transformation. How has the conversation around AI shifted in boardrooms compared to even two or three years ago and are leaders asking the right questions yet?
There was a very clear “before and after” moment with the launch of ChatGPT in November 2022. Before that, AI discussions rarely reached the boardroom and were typically driven by individuals exploring specific use cases such as credit scoring, fraud detection, or document handling in areas like insurance claims and KYC. ChatGPT changed that overnight. For the first time, people could directly interact with a generative AI model, and its free availability led to rapid, widespread adoption – in their private lives, but very fast also in their professional lives. Within weeks, executives could no longer ignore AI’s potential as a central transformation driver.
In the early stages, boardroom conversations were often shaped by instinct. Some leaders fully embraced the opportunity, while others focused on the risks. Since then, the discussion has become more grounded. Leadership teams across banks and insurers now recognise the need for a clear AI strategy aligned with business ambitions. At the same time, questions around data readiness, process maturity, legacy integration, and workforce adaptation have moved to the forefront.
Execution, however, remains the real challenge. Many institutions are still operating within their existing models, using AI to improve and automate current ways of working rather than fundamentally rethinking them. There is often a strong focus on personal productivity tools, with slower progress on enterprise-wide transformation. To realise the full value of AI, that mindset will need to shift.
Q2. There’s no shortage of AI experimentation happening across banking, insurance, and wealth management but you help institutions move beyond experimentation. What separates the firms that are successfully embedding AI from those that remain stuck in pilot mode?
Data has always been central to AI; and for good reason. High-quality, accessible data, managed securely, remains the lifeblood of AI value realisation. As we move towards agentic AI – breaking down data silos and building a company-wide data layer, enriched with ontologies and taxonomies that provide business context across domains – becomes even more critical.
AI governance is a key factor in making that a reality. Organisations need to identify and structure the risks associated with AI and make conscious decisions about acceptable risks, while still enabling innovation. Companies that strike that balance and tune their governance approach to their ambition, their DNA, and their risk appetite, tend to move faster and more confidently.
A third element is persistence in delivery. Success in AI is no longer mainly about building strong models. It is about integrating AI into existing technology environments, business processes and organisational structures. The work involved in designing, implementing, testing, and getting approval for end-to-end solutions requires discipline and persistence. These are qualities not always associated with AI initiatives.
And finally, people. Attracting and retaining talent that combines AI expertise with financial services knowledge remains challenging. At the same time, as AI moves into production, the demands on the broader organisation increase. The entire workforce needs to be both willing and able to use AI tools and to use them effectively.
Q3. AI in financial services collides directly with legacy infrastructure, regulatory scrutiny, and trust-sensitive customer relationships. In your experience, which of these three presents the greatest barrier to meaningful AI adoption and why?
All three are highly relevant and financial institutions will need to address them in parallel.
Legacy infrastructure is a clear constraint. Traditional financial institutions operate complex IT environments that have often evolved organically over decades. That is not an ideal starting point. Newer players like Revolut demonstrate what is possible when you build an AI-ready architecture from the ground up. Incumbents don’t have that luxury and they need to work around existing systems, often by adding intelligence and orchestration layers on top. That doesn’t prevent AI adoption but it does make it more complex and slower.
Regulation is another large factor. Increasing scrutiny, particularly in Europe, can feel like a burden, especially when compared to other regions. But it also reflects customer expectations around security and trust. Rather than treating regulation as a box-ticking exercise, there is an opportunity to use it as a foundation for building trust with customers.
The nature of financial services is highly regulated, trust-sensitive and data-intensive. This certainly increases the challenges for AI adoption and at the same time, it also raises the bar for potential impact. That’s what makes it such an interesting space.
Q4. When you’re working with a senior leadership team, how do you help them identify where AI genuinely creates business value versus where it simply introduces new risk? What does that decision-making framework look like in practice?
The starting point always needs to be the business strategy, rather than the technology. Planning the right AI journey is about asking what problems the leadership wants to solve or what ambition they want to achieve, instead of what AI is capable of doing. It’s also important to look at end-to-end value chains rather than isolated tasks, as that’s where the real impact tends to sit.
That approach helps focus on value beyond cost reduction, including improvements in quality, client experience, and scalability and potentially also more fundamental shifts in products, services and client interaction models.
From there, decisions are made using a multi-criteria framework. Business value is balanced with implementation cost and timelines, data readiness and security and the complexity of both the AI solution and its integration into existing systems. Governance and regulatory considerations also play a role, alongside the people impact. In particular, how processes and ways of working will change, and what level of adoption effort is required across the organisation.
Q5. Data and process readiness are often cited as prerequisites for successful AI implementation, yet many institutions underestimate what that actually requires. What do organisations consistently get wrong when it comes to preparing their foundations for AI?
I wouldn’t say organisations consistently get it wrong but rather that the target keeps evolving. AI technology is maturing rapidly, and with it, the requirements for data and process foundations are expanding. What is needed to train and run traditional AI models is very different from what is required for generative AI as a productivity tool, and different yet again from what it takes to support agentic AI operating in real time.
At the same time, there is a constant tension between building solid, enterprise-wide foundations first and delivering business impact quickly. Organisations want to realise value from AI now, rather than waiting for large-scale transformation programmes to complete. That tension isn’t new but the time available to resolve it is shorter in the AI era and does create the risk of shortcuts being taken.
Q6. As a female leader operating at the intersection of AI strategy and financial services two fields historically dominated by men what has your experience been, and what would you say to women earlier in their careers who are navigating similar spaces?
Financial services and technology are still male-dominated, particularly at senior levels. That shows up in small ways, such as being the only woman in the room. And also in bigger ways, such as fewer sponsors, fewer visible role models and often having to prove yourself that bit more.
For women earlier in their careers, a few things matter. Seek out sponsors as well as mentors. These are people who actively create opportunities and support you when decisions are made.Don’t wait until you feel completely ready to take the next step; build confidence through action and make your impact visible. It’s also important to find your voice in senior settings: being clear, fact-based and outcome-focused goes a long way.
Finally, choose your environment carefully. The right organisation and leadership team will not only challenge you but also invest in you and remove barriers. Those environments do exist and they make a real difference.
Q7. Looking ahead, which AI-driven shift do you believe will have the most profound and lasting impact on how financial institutions serve their clients over the next five years and is the industry prepared for it?
A five year outlook is long term in an AI context. I would expect to see the shift from AI that assists people to AI that executes work become real. In practice, that means agentic AI embedded across end-to-end client and operational value chains. As models continue to improve, potentially accelerated by advances such as quantum computing, this becomes increasingly feasible.
For clients, this could materially change the experience: faster onboarding and servicing, more consistent advice and far more proactive interactions. Whether around fraud prevention, protection gaps, liquidity or larger life events. For institutions, it offers a step-change in cost, speed and quality – provided it is implemented with the right guardrails. Is the industry prepared? Some are, but many are not. At least not yet. The challenge is data quality and governance, process discipline and the ability of people and organisations to adapt. Those that combine ambition with that level of discipline will move beyond incremental gains to genuinely differentiated client value propositions.
