In February 2026, Commonwealth Bank posted a half-year profit of $5.45 billion. Weeks later, it cut 300 jobs. The Finance Sector Union called it "totally unacceptable." CBA called it preparation for an AI-powered future.
It wasn't the first time. In mid-2025, CBA tried to replace 45 customer service staff with an AI voice bot — then reversed the decision after the FSU took the bank to a workplace tribunal. The union proved call volumes were rising, not falling, and CBA admitted it hadn't "adequately considered all relevant business considerations."
These aren't isolated stories. They're early signals of a shift that will touch every corner of Australian finance.
A million workers, one question
About 519,400 Australians work in financial and insurance services, according to ABS data from December 2025. But finance jobs extend well beyond banks and insurers — add in accountants, bookkeepers, payroll staff, and real estate agents across every industry, and the number climbs past one million.
Our data covers 18 finance-related occupations. Their average AI exposure score is 6.3 out of 10. Nine of those occupations score 6.5 or above — a combined 404,800 workers in the highest risk band.
That makes finance one of the most AI-exposed sectors in the Australian economy. You can see where your occupation sits on our full rankings.
Which roles face the highest risk?
The pattern is clear: the more routine and data-heavy the work, the higher the AI exposure.
Highest risk (AI score 7.0+):
- Accounting clerks — 143,500 workers, AI score 7.2
- Bookkeepers — 90,400 workers, AI score 7.1
- Financial dealers — 20,100 workers, AI score 7.1
- Insurance, money market and statistical clerks — 40,500 workers, AI score 7.0
High risk (AI score 6.5–6.9):
- Payroll clerks — 49,500 workers, AI score 6.9
- Credit and loans officers — 38,200 workers, AI score 6.9
- Insurance agents — 16,500 workers, AI score 6.5
These roles share common traits: processing transactions, reconciling data, generating standard reports, and following rule-based procedures. Generative AI and agentic systems can already do most of this faster and at lower cost.
The Jobs and Skills Australia GenAI Capacity Study found that clerical tasks — many of which previous waves of automation left untouched — are now directly in the path of generative AI. About 70% of data-entry tasks are considered exposed, and roughly 36% of accounting tasks face automation risk.
The accountant paradox
Here's the strange bit. Accountants score 6.0 for AI exposure and Jobs and Skills Australia rates them as being in "Shortage." There are 225,100 accountants in Australia, and employers can't find enough of them — even as AI takes over more of what accountants traditionally do.
The explanation? The role is splitting in two. Compliance, data entry, and standard tax returns are being automated. But advisory work — interpreting numbers, guiding business decisions, navigating complex tax structures — is growing. Employers need more accountants, just not to do the same things.
Check our accountants occupation page for the full breakdown, or compare it with other finance roles using our comparison tool.
The moderate middle
Not every finance job is in the danger zone. Roles that involve relationship management, complex judgement, or strategic decision-making score lower:
- Financial investment advisers and managers — 64,300 workers, AI score 5.6
- Finance managers — 77,400 workers, AI score 5.5
- Financial brokers — 52,400 workers, AI score 5.5
- Bank workers — 36,400 workers, AI score 5.8
These roles still face change — AI will handle more of the analysis and paperwork — but the human elements of client trust, negotiation, and regulatory judgement are harder to automate.
What the Big Four banks are doing
Australia's major banks aren't waiting.
CBA launched a $90 million "Future Workforce Program" in February 2026 — a three-year initiative to map employee skills, build career pathways, and provide targeted retraining. Over 30,000 staff have already received AI-focused training, and about 5,000 employees moved into new internal roles in the past year. The FSU is pushing for these protections to be written into the next CBA enterprise agreement rather than left to management discretion.
Westpac deployed Microsoft 365 Copilot to its entire 35,000-strong workforce — the largest Copilot rollout in Asia Pacific financial services. After a 15,000-employee pilot, the bank's Chief Data, Digital and AI Officer said "AI will completely reshape how we work."
NAB has eliminated 410 positions and relocated 127 jobs offshore, though experts note that of the Big Four, only CBA has explicitly cited AI as the reason for cuts.
The FSU reports that 36% of Australian finance workers now use AI often — nearly double the 19% in 2024. The share who never use AI dropped from 24% to just 13%. And an internal survey found 72% of CBA staff are concerned about job security.
The global picture hits home
This isn't just an Australian story. Bloomberg Intelligence projects that global banks will cut up to 200,000 jobs in the next three to five years. Citi's own research suggests 54% of banking jobs have high automation potential — the highest of any sector studied. HSBC is reportedly weighing 20,000 cuts, about 10% of its global workforce.
A separate Morgan Stanley analysis found more than 200,000 European banking jobs could vanish by 2030 as lenders invest in AI and close branches.
For Australia, PwC's AI Jobs Barometer found financial and insurance services leads all industries in AI skills demand. In 2024, 11.8% of finance job postings required AI skills — roughly three times the rate of most other sectors.
The message is consistent: the jobs aren't all disappearing, but they're changing fast. The premium is shifting from knowing how to process information to knowing what to do with it.
What this means for finance workers
Assistant Minister for Productivity Andrew Leigh put it bluntly in a recent speech: the old assumption that a degree protects you from technological disruption no longer holds. AI threatens to automate non-routine cognitive work — drafting, analysing, coding — that was once considered safe.
For the million-plus Australians working in finance, the data suggests a few things:
If your work is mostly processing and data entry — reconciling ledgers, keying invoices, generating standard reports — the risk is real and the timeline is short. These tasks are already being automated at CBA, Westpac, and elsewhere.
If your work involves client relationships, judgement, or advisory services — financial planning, complex tax work, business lending decisions — AI will change your tools but is less likely to replace your role. The JSA study found augmentation consistently outweighs automation in these areas.
If you're early in your career — building AI literacy is no longer optional. Workers with AI skills command a significant wage premium, and degree requirements in AI-exposed roles are already dropping.
The shift from processing to advising, from data entry to data interpretation, is already underway. The finance workers who adapt to that shift will find plenty of demand. The ones whose roles are defined by tasks a machine can replicate won't.
Want to see where your specific role sits? Check your occupation's AI risk score or explore all 358 occupations in our rankings.