From admin to adviser: payroll’s AI pivot

IN Partnership with

AI has already improved fraud detection by 90% – and processing errors are down 78%

PAYROLL has undergone a seismic transformation. What was once a back-office, administrative function is now a core strategic business partner – one that blends legislative mastery, data analytics and automation to drive organisational value. And as artificial intelligence redefines the role of payroll professionals, industry leaders are urging organisations to adopt early, innovate responsibly and invest in the right skills to stay ahead. Speaking to HRD, Ross Heron, CEO of the Australian Payroll Association, explains that the next wave of payroll innovation will mark a shift from retrospective reporting to predictive insight. “Payroll is about to shift from hindsight to foresight,” Heron says. “Intelligent AI will constantly map legislative updates against live pay data, flagging problems before they become liabilities, giving payroll teams the insight to fix issues on the spot.” And that future is already arriving. Liviu Marciu, CTO at Frontier Software, points to recent industry data showing major efficiency improvements thanks to AI – namely that fraud detection has improved by 90%, employee query response times have dropped by the same margin, and processing errors are down 78%.

Established in 1983, Frontier Software is an acknowledged leader in the provision of a global HR & Payroll software solution. Our software is called ichris (International Comprehensive Human Resource Integrated Software) and sets the benchmark for human capital management software. Frontier Software provides HCM solutions and services to over 1,700 customers in 26 countries, from small to very large enterprises. Our software is used to provide an efficient payroll service from our global offices in Australia, India, Malaysia, New Zealand, the Philippines, Singapore and the UK. Frontier Software ensures proven long-term growth via prudent investment in its people and software products and services. A perfect solution.

Find out more

“Ensuring AI systems are transparent and regularly updated to reflect regulatory changes is essential to balancing innovation with risk”

Nick Southcombe, Frontier Software

“Only a small number of organisations currently use AI payroll technologies,” Marciu says, “but the rate of adoption is increasing fast.” That small number is notable in Australia: research from the Australian Payroll Association found that just 4% of payroll professionals are currently deploying AI, while 57% simply don’t understand how it works or how to implement it effectively. However, the tide seems to be turning this year with HR leaders looking to invest more in new technologies in the months to come. Matthew Westwood-Hill, chief information security officer at Frontier Software, foresees these investments and innovations playing out in a very particular way. “I think we’ll see more automated data entry and reconciliation,” he tells HRD. “That means reducing manual processing time and error rates through intelligent document recognition and natural language processing.” Westwood-Hill adds that payroll anomaly detection in machine learning models and a rise in chatbots and virtual assistants will also drive dramatic efficiency gains in the months ahead.

Nick Southcombe, CEO of Frontier Software, expects a similarly dramatic impact. He says AI is already transforming payroll operations by automating high-volume, repetitive tasks and improving compliance accuracy. Tasks like extracting timesheet or contract details have achieved up to 80% efficiency gains. AI-driven compliance tools are now flagging issues such as incorrect leave accruals in real time, allowing teams to correct problems without delay. These early results are driving the next wave of innovation. AI-generated reporting is enabling faster, data-informed decision-making, while new capabilities such as predictive workforce planning, natural language processing for employee queries and automated superannuation calculations are close behind. Southcombe sees these developments “elevating the strategic role of payroll” beyond its traditional administrative limits.

Heron believes the key to this evolution is collaboration, particularly between humans and machines. He offers a concrete example: Beryl, the Australian Payroll Association’s own AI assistant. Designed to be a “co-pilot, not a replacement”, Beryl helps practitioners answer complex pay queries instantly.

“When people see the tech as a co-pilot, you unlock innovation and strengthen risk control,” Heron explains. And industry data seems to chime with the experts’ sentiments here, especially for the HR world at large. Research from The Access Group found that 68% of HR professionals are now using AI at work, adding that it saves them around three hours a week, with 98% revealing AI has had a positive impact on their workplace. That message to embrace technology but do it wisely resonates with Marciu as well. While the benefits of AI in payroll are clear, he warns that the risks can’t be ignored. “The most significant risks emerge at the intersection of compliance obligations, data privacy and cybersecurity vulnerabilities,” he tells HRD. These include algorithmic bias, misinterpretation of evolving regulations and exposure of sensitive employee data. Marciu stresses that governance must evolve just as rapidly as the tools themselves. “Success in AI adoption isn’t measured solely by technological sophistication – it’s measured by improved accuracy, enhanced compliance, reduced risk and increased employee satisfaction.”

“I would encourage payroll professionals to add three strings to their bow.AI fluency – grasp how the algorithms reach their answers. Data literacy – read the dashboards, question the anomalies ... And deep legislative mastery”

Ross Heron, Australian Payroll Association

Southcombe agrees that risk management must be baked into every phase of AI implementation. He cites the Privacy Act 1988 and the Notifiable Data Breaches scheme as critical frameworks that payroll teams must understand, especially when systems handle sensitive identifiers such as tax file numbers and bank details. Weak encryption or outdated vendor systems can expose businesses to ransomware or accidental breaches. For Westwood-Hill, these cybersecurity concerns are part and parcel of the daily grind. He says that while AI offers innumerable benefits, it’s essential to have a risk strategy in place from the off. “If AI models ingest sensitive payroll data such as names, TFNs or salary, attackers may extract or infer personal data via prompt injection or model inversion,” he adds. “[What’s more], malicious actors may intentionally poison training data if exposed, skewing model behaviour. “AI decisions, especially in areas like awards interpretation or wage compliance, may [also] lack explainability. This creates a regulatory and ethical concern, particularly under Australia’s Privacy Act and to-be-introduced AI accountability standards.”

To mitigate these risks, Southcombe believes organisations should implement strong governance frameworks, conduct regular audits and partner with vendors certified to standards like ISO 27001. Compliance, too, presents a challenge, especially when AI misinterprets complex award conditions. Here, Southcombe refers to high-profile underpayment cases, including Woolworths’ $315 million issue, as cautionary tales. “Ensuring AI systems are transparent and regularly updated to reflect regulatory changes is essential to balancing innovation with risk,” he explains. Yet for all the risks, AI adoption in payroll is no longer optional. As Heron puts it, start small with AI, run tightly controlled pilots, and keep your team in the review loop so they help shape the rules. By involving staff from the beginning, organisations can demystify the technology and strengthen both risk management and team engagement.

57% of payroll pros don’t understand how AI works

Tech as a co-pilot

Published 15 Sep 2025

Cautionary tales of AI misuse

Southcombe recalls a successful implementation focused on automating award interpretation and STP compliance for small to medium-sized businesses. Challenges included training the AI to interpret complex award structures under the Fair Work Act 2009, integrating it with legacy data systems and earning user trust. Through extensive testing and iterative updates, the platform achieved a 50–70% reduction in payroll processing time, improved STP accuracy and allowed teams to focus on higher-value work like cost analysis and planning. “It demonstrated AI’s potential to enhance both efficiency and compliance,” he says.

50–70% reduction in payroll processing time

So, what happens to payroll professionals in this new world? Heron believes the human role is far from obsolete – instead it’s changing. “I would encourage payroll professionals to focus on adding three strings to their bow,” he says. “AI fluency – grasp how the algorithms reach their answers. Data literacy – read the dashboards, question the anomalies, turn the numbers into a story. And deep legislative mastery so you can push back when the machine misses a nuance.” When combined with close collaboration across HR, finance and IT, Heron believes these skills will make payroll not just compliant but indispensable. “The way payroll professionals undertake their tasks will transform to benefit both employees and organisations. This transformation will allow them to spend more time on higher-value strategic work and improve employee engagement.”

Southcombe urges professionals to experiment with AI regularly and adopt what he calls an “AI-first mindset”. In practical terms, that means approaching every task as if it’s the first time and asking, “Could AI do this better?” At first, it may feel awkward, he says, but over time it becomes second nature. Westwood-Hill concurs, noting that data literacy roles will advance, especially those that revolve around understanding AI interpretations of payroll inputs and outputs. “Roles like payroll analysts will evolve into AI data quality governance or oversight, validating model outputs,” he adds. “Manual data entry roles may shrink, but exception handling and compliance audit functions will grow.” Southcombe emphasises the same shift – namely that roles focused on repetitive tasks such data entry and reconciliation are likely to fade. In fact, McKinsey estimates that up to 30% of routine work could be automated by 2030. In contrast, strategic roles in compliance oversight, analytics and AI system coordination will grow, always with human insight leading the way. “Professionals who can manage AI tools and ensure regulatory alignment will remain indispensable,” Southcombe says.

End of repetitive roles?

Real-world deployments already show the promise of this approach. Marciu highlights an AI-driven fraud detection system that uses machine learning to identify anomalies by comparing new transactions against historical patterns. Payroll administrators can then focus their attention on the highest-risk items, significantly reducing false positives. “This AI-driven approach transforms what was traditionally an error-prone and resource-intensive manual process into a streamlined, data-informed workflow,” Marciu adds. Importantly, the system retains a ‘human-in-the-loop’ model – AI flags the issues, but people validate them. It’s an architecture that balances efficiency with accountability.

Biggest risks for payroll in 2025

ASIA

News

SPECIALISATION

Events

Best in HR

Resources

Companies People Newsletter

About us Authors Privacy Policy Conditions of use Contact us RSS

Technology upgrades – 24.5%

Automation of tasks – 18.1%

Poor or incomplete data coming into the payroll office – 32.7%

Data security – 14%

Not finding the right payroll talent – 13.2%

Poor integration between systems – 23.6%

Current payroll challenges

Payroll technology and process – 22.3%

Keeping up to date with understanding legislation – 19.9%

Award and EBA interpretation – 15.6%

Incomplete or inaccurate data – 15.3%

News

Specialisation

EVents

Best in HR

Resources

ASIA

News

Specialisation

EVents

Best in HR

Resources

ASIA

Companies People Newsletter

About us Authors Privacy Policy Conditions of use Contact us RSS

Source: Australian Payroll Association

Companies People Newsletter

About us Authors Privacy Policy Conditions of use Contact us RSS