The Finance Talent Shift That Has to Happen Before Any AI Tool Lands

Roland Berger's readiness data on ASEAN GBS shows that thirty-seven percent of organisations with an AI vision have actually started training their people for it. That number matters because most AI-in-Finance timelines quietly fail at the training stage, and they fail because of the team well before the technology is ever the problem.

A Finance function commissions an AI build, and it works, but somewhere in the rollout it becomes clear the people running the process were never re-skilled for the version of the job the AI creates. The AP Analyst hired in 2023 was brought into process invoices accurately and quickly. By 2027 that same person processes almost nothing and instead spends their time judging the exceptions a model flags and reading the anomaly pattern that signals something is wrong before the number does. The job title stays the same while the work behind it changes completely.

What's already being mapped

This is not a problem any single Finance function is facing in isolation. Malaysia's Human Resources Ministry told Parliament in June that 697,000 existing jobs across ten priority sectors are highly vulnerable to disruption within three to five years unless workers reskill, and that 120 new roles are expected to grow in importance over the same period. Global Business Services is one of those ten sectors. The government has already identified where the shift lands and is now building the reskilling response around it, which is the same sequence an individual Finance function has to run at its own scale.

TalentCorp's MyMahir Future Skills Talent Council is the delivery mechanism for that national effort, and it is already mapping the finance-specific roles this shift demands. At a recent event in Penang on the future of GBS careers, Chang Te Sheng walked an audience of professionals and students through this exact transition, showing how the Financial Analyst becomes the AI-Augmented Finance Business Partner and the AP Analyst becomes the Process Automation Lead. It is the same pattern the Council is tracking, made visible in a room instead of a report.

Organisations that hire and develop against that direction will outpace the ones still redeploying yesterday's job descriptions into tomorrow's work.

What it asks of the people already in the seat

Processing an invoice correctly is a skill you either have, or you don't, and most people can be taught it in a week. Judging whether an exception is a data error, a process gap, or the early signal of fraud is a harder skill that takes longer to build, feels less comfortable to practise, and exposes people to being wrong in front of colleagues in a way that following a checklist never did.

Some Finance professionals will work through that discomfort and come out more valuable than they were. Others will wait for it to feel easier before they engage, and by then the decision about their role will have been made without them.

This is more than a talent problem for HR to handle. The CFO has to own it as a sequencing problem, because in most organisations the sequence is currently backwards. Capability building gets treated as a follow-up task once the AI tool goes live, tucked into the change management plan and usually underfunded next to the technology spend it is meant to support.

Why the order matters more than the intent

That sequence must be reversed. The retraining, the role redesign, and the honest conversation about which parts of a job are disappearing, and which are about to matter far more all have to be substantially done before the tool lands. Change management best practice is not what drives this. The driver is the twelve-month ROI timeline every CFO is now held to, which leaves no room for a team still learning to work with the tool six months after go-live.

If you are mapping an AI deployment in Finance for the next two quarters, the honest question to put to your own team is a simple one. Which three roles change shape first, and has anyone started teaching the judgement calls those roles will need before the software that surfaces them arrives?

The Finance talent shift is already mapped. What remains uncertain is whether your hiring and training have moved with it.

---

AGOS Asia is an AI-first digital GBS transformation partner operating across ASEAN. We work with finance and shared services leaders to rethink, redesign, and rewire their operating models for the AI era. AGOS GBS Summit 2026, the ninth edition, takes place on 10 September at Sheraton Petaling Jaya.

Register for AGOS GBS Summit 2026