Xero study says AI unlikely to cut Australian jobs

Xero has published research showing that most Australian accounting and bookkeeping practices do not expect artificial intelligence to cut jobs over the next year. Only one in eight practices expect headcount to fall.

Based on responses from 524 independent senior accountants and bookkeepers in Australia, the data instead points to changes in hiring, pricing and service mix as firms adapt to wider AI use.

Across the market, 88% of practices said they did not expect AI to reduce team numbers in the next year. While concerns about job losses have surrounded AI in professional services, the findings suggest many firms are using the technology to reshape roles rather than remove them.

Some 38% of respondents said they expected to restructure roles and teams around new ways of working linked to AI. A larger share, 63%, said they were changing how they hire, with greater emphasis on soft skills and relationship management.

The results also draw a distinction between firms that have tried AI and those that use it regularly. Although 85% of practices said they had experimented with the technology, only 52% said they used it day to day. Among top-performing firms, that figure rose to 69%.

Profit gap

The most profitable practices in the study were earning 2.8 times more than lower-margin firms. They were also more likely to have embedded AI into daily workflows, shifted time into advisory work and adopted pricing models that move away from billing by the hour.

Top-performing practices reported saving 11 hours a week through AI, compared with eight hours across all firms. They also estimated annual savings of AUD $244,000, versus AUD $160,000 for the wider sample.

The study found firms were directing time freed up by AI into several areas. Across all respondents, 44% said they were using that capacity for higher-value advisory work, while 40% pointed to innovation and new workflows, 39% to training and upskilling, and 36% to business development.

Charlie Sheppeard, Executive General Manager, Strategy & Operations, Australia and New Zealand at Xero, said the difference lay in how firms were applying the technology, rather than whether they had simply tested it.

"Accounting is regularly named among the professions most exposed to AI disruption. Our data shows practices that are making deliberate choices about where AI fits, how their people use it, and how they price the value they create are seeing the strongest growth," said Charlie Sheppeard, Executive General Manager, Strategy & Operations, Australia and New Zealand at Xero.

The research also linked more routine AI use with better-defined internal processes. Among practices using AI in day-to-day work, 94% rated their core business processes as well documented, compared with 40% of those not using AI.

Advisory shift

Just over half of Australian practices in the survey, 52%, said they currently offer advisory services. Of those firms, 48% said advisory had lifted revenue by more than 30%. Respondents also ranked it among their highest-earning service lines, alongside tax.

The figures suggest firms are using AI-driven efficiency to support a broader shift in how they make money. Rather than relying chiefly on compliance tasks and time-based billing, more practices appear to be moving towards advice-led work and fixed or value-based pricing.

On average, nearly seven in 10 client engagements, or 69%, were billed on non-hourly terms. Among practices using value-based pricing, 43% said it had improved profitability, while 49% said it made client expectations easier to manage.

Top-performing firms were also charging more for advisory work. According to the research, they charged 34% more for advisory services than lower-margin practices.

Andrew Erkins, Founder of Digit, a bookkeeping, advisory and payroll firm, said the key issue for the sector was how professional roles would evolve as AI becomes more common.

"For accountants and bookkeepers, the real question isn't about getting replaced; it's what our role will become. That means embracing AI and supporting our teams and clients on how to use it. They should be encouraged to learn new tools and given guidance and security guardrails to build their own. The goal isn't to become coders; it's to help our industry define the shape of a problem and see how technology could solve it differently from the way we have in the past," said Erkins.