Australian CFOs believe AI in finance will be transformative, but five blockers stand in the way. Learn how to overcome these and scale your agentic AI today.
Luke Bebbington
Industry Advisor, Financial Services
Workday APAC
At a recent Workday event, I wasn't surprised to hear Australian CFOs and senior finance leaders say they are under increasing expectation to use AI to lift productivity, improve insight and help the business move faster.
But I was intrigued to see that, when asked about their confidence to actually deploy AI in finance, there was a striking gap in the results.
On average, our participants at Finance Think Tank events rated the expected impact of AI on finance at 4.1 out of 5. But they rated their professional AI capability at just over 2 out of 5.
Even more telling, most rated their personal use of AI higher than their workplace use. Turns out finance leaders are experimenting with AI. Just not always at work!
It got me wondering: what’s really holding CFOs back from moving beyond pilots and productivity hacks to embrace the possibilities of AI?
Especially when it comes to agentic AI, which can monitor systems, interpret real-time conditions, and initiate tasks across connected applications.
Report
A clear pattern emerged from our Finance Think Tank events. Local finance leaders are looking at multiple areas where AI has great potential to transform their function, including financial reporting, accounts payable and FP&A.
In finance, trust matters most and there is no acceptable margin of error. When payroll runs, the books close, or compliance is on the line, almost right is wrong.
For CFOs under constant pressure to deliver faster and more granular reporting without adding headcount, general ledger and financial close is often an obvious starting point for AI enablement.
But I couldn’t help noticing the contrast between this ambition and the tools finance teams are actually using.
Despite all the digital potential in the world, our participants reported that Excel and offline models remain a significant part of their toolkit.
This reflects the reality of finance tech infrastructure. Where legacy systems still hold sway, spreadsheets are likely the only option.
What this means is that individual-level innovation is running ahead of enterprise adoption.
Organisations are already using generative AI in finance to move faster, but often outside core systems and formal controls.
Our participants said they’re making much more use of public LLMs, like ChatGPT, Claude and Gemini, than ERP-based AI tools.
It all points to Australian finance organisations being in a paradoxical situation.
The team is pretty sure where AI can deliver the most value. But the conditions needed to realise that value at scale are still catching up.
In my conversations with industry experts, I asked why this was the case. They were clear that the main blockers aren’t technical – they’re organisational.
Their observations about the top blockers tie in with what the CFOs at our Finance Think Tanks are seeing on the ground:
1. Controls, governance and auditability: In finance, trust matters most and there is no acceptable margin of error. When payroll runs, the books close, or compliance is on the line, almost right is wrong. I believe this is the biggest reason that CFOs in Australia are hesitant to adopt agentic AI in their organisations.
2. No clear owner for AI transformation. Agentic AI introduces a different level of autonomy and accountability than traditional automation. Middle management often hesitates to take ownership of AI agents that feel risky or unfamiliar, particularly when benefits span multiple functions. Without clear sponsorship, transformation stalls.
3. Agentic AI skills: Many finance teams have become expert consumers of existing technology, but they lack the skills needed to reimagine how work gets done. This is more than just understanding how to prompt or build an agent – it's about innovating from the ground up, instead of simply improving current processes.
4. Shadow AI usage. As our Finance Think Tank participants told us, employees already use generative AI to write, summarise and analyse – often quietly. Having the right balance of governance and trust in your people is critical, so they can experiment in a safe space that doesn't compromise the integrity of the finance function.
5. Difficulty proving the business case. Finance leaders know better than anyone that transformation needs a baseline. Yet many finance functions lack solid metrics around effort, cycle time and decision quality, making it hard to quantify the upside of AI-driven change.
Many Australian leaders maintain a cautious approach – prioritising risk management and operational resilience over the immediate integration of frontier AI tools.
Agentic AI is fundamentally different from generative AI add-ons. Agentic systems don’t just respond to prompts. They can orchestrate steps across close, reporting, planning and controls, escalating exceptions, gathering context and learning over time.
According to McKinsey, agentic AI is a key area where finance teams are delivering the most value with AI.
McKinsey gives the example of a global biotech that replaced manual invoice-to-contract compliance with an agentic AI system that promptly identified previously unknown contract leakage of US$40 million!
Agentic AI in finance can confer hugely powerful analysis and efficiency capabilities. But these systems only work if finance organisations have:
Trusted, unified data foundations
Embedded governance and controls
Clear decision rights and accountability
People with skills and experience using AI agents
People confident enough to redesign work, not just augment tasks
Without these foundations, finance teams risk automating one bottleneck only to create another. And this is likely why finance is falling behind the rest of the organisation.
The Reserve Bank of Australia’s November 2025 Bulletin noted that while Australian firms have invested heavily in foundational digital infrastructure, AI adoption remains largely piecemeal and limited to basic digital assistants.
The report reveals that nearly 40% of surveyed firms report only minimal AI usage, as many Australian leaders maintain a cautious approach – prioritising risk management and operational resilience over the immediate integration of frontier AI tools.
According to the Finance Evolution: Thriving in the Next Decade report from Chartered Accountants ANZ, finance teams have a five-year window to evolve their role – or risk becoming irrelevant to business strategy.
The research highlights growing gaps in digital, data and sustainability capabilities, alongside persistent concerns about finance being seen as a cost centre rather than a value driver.
Finance leaders need an understanding of what’s possible – and the confidence and determination to make building the foundations and reskilling the team's top priorities.
AI and analytics are essential. But they’ll only deliver value if finance teams learn new technology skills. In other words, the blocker isn’t AI itself. It’s whether finance functions are ready to pair new technology with new ways of working.
Finance teams have to develop the digital infrastructure and human capabilities required to get serious about agentic AI. Instead of tinkering with pilots or experimenting with generative AI, CFOs need to:
Fix the basics, prioritising investment that addresses data quality and integration
Build capability inside the team, combining data literacy with confidence to redesign processes
Link agentic AI ambition directly to finance outcomes, starting with low stakes areas such as running forecasts in parallel, or accounts payable processing
Start small, but intentionally, using early wins to rebuild innovation muscle
Make agentic AI governance explicit, so experimentation happens safely – not secretly
Australian CFOs may not think they have the professional capability to leap into agentic AI adoption. But technical competence is not what’s missing.
Instead, finance leaders need an understanding of what’s possible – and the confidence and determination to make building the foundations and reskilling the team's top priorities.
Report
Discover how agentic AI can add value to your finance function, based on real experiences from Australian leaders like you.