AI IN FINANCE
74% of finance leaders prioritise AI, yet ambition is outpacing adoption
Finance leaders rank AI integration as a top priority for the next 12 months, alongside stronger financial controls and resilience against market volatility.
90% expect to increase AI investment in finance over the next year, but 35% don't plan to extend AI into more workflows. What’s holding them back?
To find out more about the finance divisions’ experiences and expectations of AI within their function, we commissioned Forrester Consulting to survey 1,279 senior finance leaders across 11 markets.
One in three finance leaders aren’t expanding AI this year. Here’s why.
say fragmented, inconsistent data is a key challenge to scaling AI across finance workflows
cite limited ability to configure, extend, or customise vendor-provided AI capabilities
cite their teams' limited experience in running AI-enabled finance processes
Deep dive into our insights on AI in finance
The geography of AI in finance: why North American teams are pulling ahead
37% of North American finance workflows already involve multi-step AI execution. European teams, it seems, have a lot of catching up to do.
The build vs buy dilemma: Why most finance teams are hedging their bets on a hybrid model
51% are moving towards a hybrid model. The question is, when do you draw the line betweeen internal expertise and outsourcing?
AI sped up the workflows, but human oversight became the bottleneck
Productivity went up. Processes weren’t faster. 84% still require manual work to complete workflows. Finance teams are starting to understand why.
The AI ROI question CFOs can't answer (yet)
74% of finance leaders say AI is a top priority. Yet 40% are still stuck building the business case, and struggle with justification beyond short-term productivity gains.
Building an AI-Ready Finance Function
A Forrester report commissioned by Airwallex.
Finance teams aren’t hesitating on AI, but they’re being held back by their foundations
Current infrastructure can’t deliver what finance leaders expect
When finance leaders identify what's blocking them from scaling AI, the top five technology challenges point to the same underlying problem: infrastructure that wasn't built to move data cleanly across entities, regions, and functions.
- 74% have already moved toward a single platform or a smaller number of core platforms to support their finance workflows.
- Yet, 84% still require manual intervention to complete their finance workflows.
- Fragmented, siloed information across entities and regions makes it harder to scale AI beyond simpler use cases like reconciliation, reporting, and basic AP flows.
- Consolidating platforms reduces complexity, but it doesn’t resolve the issue if data cannot be moved cleanly across systems.
Orchestration is the missing layer, point solutions are losing ground
The value of AI comes from how effectively workflows can be orchestrated across systems. Finance leaders are clear that they need an orchestration layer to create a foundation for AI to work across the full finance function.
- 66% say orchestration across the broader finance ecosystem is their top vendor criterion.
- 65% want flexible, platform-based architectures that support incremental adoption.
Current infrastructure can’t deliver what finance leaders expect
When finance leaders identify what's blocking them from scaling AI, the top five technology challenges point to the same underlying problem: infrastructure that wasn't built to move data cleanly across entities, regions, and functions.
- 74% have already moved toward a single platform or a smaller number of core platforms to support their finance workflows.
- Yet, 84% still require manual intervention to complete their finance workflows.
- Fragmented, siloed information across entities and regions makes it harder to scale AI beyond simpler use cases like reconciliation, reporting, and basic AP flows.
- Consolidating platforms reduces complexity, but it doesn’t resolve the issue if data cannot be moved cleanly across systems.
Orchestration is the missing layer, point solutions are losing ground
The value of AI comes from how effectively workflows can be orchestrated across systems. Finance leaders are clear that they need an orchestration layer to create a foundation for AI to work across the full finance function.
- 66% say orchestration across the broader finance ecosystem is their top vendor criterion.
- 65% want flexible, platform-based architectures that support incremental adoption.
“Many tools are adding AI capabilities, but they operate within their own walled gardens. You can use AI inside a single application, but it doesn’t communicate with AI in other systems. That’s where native integrations fall short. If you need human review for exceptions, context across multiple systems, or logic that spans workflows, you need an orchestration layer.”
Chief Financial Officer
Technology (SaaS), US
“For me, the most important factor is how well these tools integrate across the finance ecosystem. We recently implemented a treasury management system, and it’s been very effective because it connects seamlessly with our other systems. That flow across platforms has allowed us to keep the treasury team lean and automate a significant portion of the process.”
Senior Finance Director
Technology (Travel), UK
“Progress toward autonomous finance needs to be assessed by function rather than treating finance as a single block. In transactional areas such as accounts payable and receivable, they are far easier to automate because the rules are clear and outcomes are largely binary. Strategic finance remains less advanced. The work is inherently more complex, relying on judgement, business context, and well-integrated data.”
Head of Commercial Finance
Ecommerce , Australia
“Finance automation is advancing faster than audit practices can keep up In finance, correctness isn’t optional, it’s critical. I see it as my responsibility to ensure the numbers are right. I think we’ll get closer to full automation over time, but the broader ecosystem needs to catch up, particularly auditors.”
Senior Finance Director
Technology (Travel), UK
“For AI to be effective in this space, it needs to learn the nuances of the business, what the key drivers are, what matters, and what “good” looks like. That takes time, iteration, and continuous feedback. Until that context is built, strategic finance won’t become autonomous at the same pace as accounting, even as AI becomes a core tool in how the work gets done.”
Chief Financial Officer
Technology (SaaS), US
What finance leaders want: Real-time finance intelligence
Faster insight-to-action
57% say the main reason to adopt AI is to improve real-time forecasting and scenario planning, alongside the end-to-end automation of routine workflows. AI is moving up the value chain from efficiency gains to real‑time insights and decision support.
AI as a co-pilot
Only 23% are aiming for AI agents that change decisions on their own. What finance leaders are building toward is AI as co-pilot: better foresight and more reliable execution, with humans overseeing strategy.
A unified platform
More than 50% are prioritising diagnostic and predictive capabilities to inform financial decisions. That kind of intelligence requires a foundation where data moves consistently across entities, regions, and functions from the start.
Connected infrastructure is what closes the gap
A strong foundation has to be built from the ground up. Airwallex is built AI-native and API-first. Data is orchestrated across payments, treasury, and spend management by design. We’ve spent years securing licences, building payment rails, and connecting global schemes to move money across borders. This foundation gives AI visibility into the complete flow of funds, reducing fragmentation and enabling consistent execution at scale.
AI adoption rarely moves at the same pace as internal AI capability-building. Our platform is designed to run complex finance workflows without requiring deep technical expertise on your team. Kai, our AI assistant, help users set up their account, automate workflows, and answer data questions using natural language.
countries where you can collect funds like a local
countries to which you can make local transfers
287B+
global payments processed annually
countries from which you can accept payments
AI execution stalls without the right people to run it
53% of finance leaders cite skills gaps or limited experience running AI-enabled finance processes as a key challenge in scaling AI. Most are still in the early stages of closing that gap.
- Only 15% are running a structured, enterprise-wide AI talent strategy that includes role redesign, certifications, and hiring.
- 14% haven't yet assessed what skill changes AI will require, let alone launched training.
- The majority are somewhere in between, with 24% building basic literacy and 30% training selected staff to develop AI-enabled workflows.
As AI transforms the finance function, leaders need to rethink tasks to be done across the stack and be open to hiring from non-traditional backgrounds. Finance teams will need to evolve toward hybrid skillsets, combining traditional financial expertise with data, AI, and strategic capabilities.
The path forward runs on a hybrid model
Rather than fully building or outsourcing AI capabilities, 51% of finance leaders are scaling AI through a hybrid model, combining in-house capabilities with external expertise. That figure is expected to rise to 57% over the next 12 months.
China, Singapore, and the Netherlands lead in the shift away from purely outsourcing to third-party providers, which risks vendor sprawl and data fragmentation.
AI execution stalls without the right people to run it
53% of finance leaders cite skills gaps or limited experience running AI-enabled finance processes as a key challenge in scaling AI. Most are still in the early stages of closing that gap.
- Only 15% are running a structured, enterprise-wide AI talent strategy that includes role redesign, certifications, and hiring.
- 14% haven't yet assessed what skill changes AI will require, let alone launched training.
- The majority are somewhere in between, with 24% building basic literacy and 30% training selected staff to develop AI-enabled workflows.
As AI transforms the finance function, leaders need to rethink tasks to be done across the stack and be open to hiring from non-traditional backgrounds. Finance teams will need to evolve toward hybrid skillsets, combining traditional financial expertise with data, AI, and strategic capabilities.
The path forward runs on a hybrid model
Rather than fully building or outsourcing AI capabilities, 51% of finance leaders are scaling AI through a hybrid model, combining in-house capabilities with external expertise. That figure is expected to rise to 57% over the next 12 months.
China, Singapore, and the Netherlands lead in the shift away from purely outsourcing to third-party providers, which risks vendor sprawl and data fragmentation.
North America is pulling ahead with 37% running multi-step AI workflows
Clean data is the real edge
In North America, 37% said AI is advanced enough to execute multiple pre-configured tasks, higher than the global average of 30%. That’s because only 47% report siloed or inconsistent data, which is cited globally as the primary barrier to scaling AI. This is lower than the 55% in APAC and 58% in EMEA.
Spending plans back up the execution lead
North America's investment intent backs up its execution lead, as 67% expect AI investment to increase by 10% or more over the next 12 months. Only 8% expect no change, one of the lowest rates of investment hesitancy across regions.
EMEA’s regulatory environment is widening the gap between markets
Compliance pressure isn't evenly distributed
Within EMEA, the UK stands out with the highest AI execution maturity across autonomous, multi-step pre-configured, and singular-action workflows. Only 49% say risk and compliance concerns are actively slowing AI deployment, compared to 54% in France.
Fragmentation is still the ceiling
A more flexible regulatory environment relative to parts of the EU has supported stronger AI adoption intent, but data fragmentation remains the binding constraint. 71% in France and Germany cite integration complexity across regions, entities, or providers as their top technology challenge in scaling AI in finance workflows, highest globally.
In APAC, workforce preparation determines which markets lead
Early movers built the capability first
In APAC, the markets doing the most to build internal AI capability are the ones moving the fastest. 27% in Singapore are running a structured, enterprise-wide AI talent strategy, the highest of any market in the study. China follows closely behind at 21%. That investment in people shows up in execution, where 18% in Singapore run AI autonomously with minimal human input, compared to 11% globally.
The talent gap shows up in execution
On the other end, New Zealand and Australia lag on workforce preparation. Only 7% and 9% respectively are running enterprise-wide talent strategies – and that gap maps directly onto slower AI execution progress.
See what finance leaders are prioritising next
Airwallex commissioned Forrester Consulting to conduct research to understand how finance leaders are navigating the shift from early adoption to the realities of scaling AI across complex environments.
A survey of 1,279 finance decision‑makers and interviews with six finance leaders across 11 markets provide insights into the barriers that emerge beyond initial deployment, and how orchestration aligns data, systems, workflows, and human judgement to advance AI in a sustainable way.
All statistics referenced above come from this study, June 2026.