AI in Finance: Making AI Work – Key Insights from London’s Finance Transformation Briefing

Yesterday, Kefron brought together finance leaders, technology specialists, and transformation experts in London for an intimate breakfast briefing on a topic dominating boardroom conversations: how to move from AI experimentation to real-world impact. As organisations accelerate digital transformation in finance, the focus is shifting from experimentation to measurable results powered by AI in finance.

The session, titled “AI, Automation and Finance Transformation: From Hype to Impact,” delivered on its promise of practical insight over vendor pitches. Split into two panels – one exploring the art of the possible, the other focused on delivery realities – the morning offered a refreshingly honest look at where finance technology trends are heading and what it actually takes to embed finance automation successfully.

The Future of AI in Finance: From Legacy Platforms to Intelligent Finance Systems

Paul Kearns from Kefron and Darren Cran from AccountsIQ opened the discussion by addressing the pressure finance leaders face: organisations are scaling faster than their systems can support, and legacy platforms simply are not fit for purpose in a modern environment shaped by AI automation in finance and increasing regulatory scrutiny.

But the urgency is not just about efficiency. Legislation is coming – particularly around e-invoicing. The UK is expected to mandate digital invoice verification by 2029, following countries like Poland and France where invoices must be submitted to government portals to combat fraud. HMRC has already launched consultations on e-invoicing reform, signalling the direction of travel for UK businesses.

For companies still running legacy systems, that timeline is tighter than it appears. As Kearns noted, realistic preparation for finance systems modernisation means starting at least two years ahead of implementation.

On the technology side, Cran outlined the evolution of finance platforms: moving from a system of record (the ledger), through a system of work (where teams execute tasks), to a system of intelligence, an AI layer that enables intelligent finance systems where professionals interact with data naturally, not just extract it.

The real shift in AI in finance is moving beyond chatbots to AI agents that perform actions autonomously. Within five years, routine tasks such as invoice coding, approvals, and journal preparation, particularly within AI in accounts payable, are expected to happen automatically, with humans managing exceptions rather than processing every transaction.

Kearns highlighted another application: analytics at scale. When processing millions of invoices across hundreds of customers, AI can surface insights such as overpayments, benchmarking opportunities, or cost anomalies. Natural-language reporting will replace static dashboards, users will simply ask questions and receive precise answers from their data.

Research from McKinsey highlights that AI could automate a significant proportion of finance activities while improving decision intelligence, reinforcing the trajectory toward AI automation in finance.

The question everyone is asking about AI for CFOs: what about jobs?

An audience member raised the concern many finance teams share: will AI replace people?

The panel’s answer reflected a practical view of AI for CFOs. AI is about productivity, not headcount reduction. Most CFOs do not want teams doubling as revenue grows – they want scalable efficiency enabled by finance automation.

In practice, AI in finance typically slows hiring growth rather than triggering layoffs. According to the World Economic Forum, AI adoption across professional services is more strongly associated with job augmentation than displacement.

However, there was also a clear warning: ignoring AI is risky. Those who embrace digital transformation in finance and learn to use intelligent systems effectively will be in a stronger position than those who resist. Every major technology wave, from industrialisation to the cloud, caused fear, but work evolved rather than disappeared. The same pattern is emerging in finance transformation.

The harder truth behind finance transformation

The second panel shifted from vision to reality. Rob Allen from DNATA and Phil Oddie from The PO Consultancy brought decades of experience delivering large-scale finance and procurement transformation programmes and they did not sugarcoat the challenges of finance transformation.

Allen shared a sobering example on a project that ran for over 10 years, went live, and disappeared within a year. The core issue was insufficient groundwork, no clear “why,” no shared goal, and a solution misaligned with broader finance systems modernisation strategy.

His lesson: sometimes the best decision in transformation is to stop.

He revealed another reality: while technology matters, long-term partnership matters more. Sustainable finance transformation depends on alignment, not just software delivery.

What actually makes finance automation projects succeed

Oddie, who has worked on complex e-invoicing implementations across multiple countries, identified the single most important success factor in finance automation initiatives and it is not the technology.

It is user adoption.

If people do not use the system, nothing else matters. Successful AI in finance and intelligent finance systems require strong change management: communication, engagement, and internal advocacy.

One of the most effective tools is change champions, trusted individuals who advocate locally. Resistance is strongest when change feels imposed centrally, but adoption improves when trusted peers validate the benefits of automation.

Oddie emphasised that engagement is not a tick-box exercise. Transparency builds trust. In cross-border implementations, cultural nuance, communication clarity, and simplicity significantly influence the success of finance transformation programmes.

Taking it Forward

The morning concluded with a clear message: finance leaders do not need to choose between modernisation and control. But successful AI in finance requires more than selecting new technology.

It demands early preparation for regulatory change such as e-invoicing. It requires honest assessment of whether existing platforms can support future growth. It means investing in change management alongside finance automation. And it depends on building genuine partnerships that support long-term finance systems modernisation.

Most importantly, it requires creating an environment where curiosity about AI in finance is encouraged, where teams experiment, learn, and adapt rather than resist.

The gap between hype and impact in AI automation in finance is not about the technology itself. It is about preparation, engagement, and strategic leadership. That is where real finance transformation begins.

We’re taking this conversation deeper in Dublin. Our next breakfast briefing will focus on implementation: how to secure executive buy-in, align competing priorities, and build the internal foundations that make finance transformation actually stick. Register here to join us.