Australian boards approving AI budgets this year have, in many cases, also approved hiring freezes in the same breath. The assumption behind both decisions is often the same: that AI is there to do the work people used to do, and headcount should shrink to match.
New research suggests that assumption may be backward, and it’s a data point Australian CIOs justifying AI spend to skeptical boards will want on hand.
The study, run by the finance platform Ramp and the workforce data firm Revelio Labs, tracked anonymized AI vendor spending against detailed employment records from 21,559 US companies. It found that firms spending heavily and consistently on AI tools weren’t shrinking. They were hiring, and hiring faster than similar firms that hadn’t yet adopted AI at scale — including at entry level, the segment most often assumed to be first in line for automation.
Contradicting data shows up
Researchers set a threshold to separate genuine AI adopters from firms dabbling with or randomly testing AI tools. The defining metric: at least $100 per employee per month in AI vendor spend, sustained for three consecutive months.
Firms clearing that bar with the highest intensity increased total headcount by roughly 10% over the following two years, compared with otherwise similar firms that hadn’t yet adopted. Firms that spent only lightly on AI tools showed no measurable change in headcount.
The research also revealed something surprising. Entry-level hiring at the highest-intensity adopters grew even faster, up around 12%. Engineering headcount expanded by more than 7%, with sales, customer service, finance, and administrative functions all posting mid-to-high single-digit gains. These are employment stages and roles often cited as negatively impacted by automation.
Notably, the effect wasn’t immediate. It barely registered in the first six months and only became substantial after 12 to 18 months, once AI use had moved from pilot to standard practice.
Where the report touches Australian enterprises
Australian enterprises are weighing AI investment against workforce decisions at a time when many boards still treat the two as a trade-off: spend more on AI, spend less on people. The Ramp-Revelio findings directly challenge that framing. In the firms studied, AI spending intensity tracked with more jobs, not fewer — but only once adoption moved beyond the trial stage.
That distinction matters in Australia’s current environment. Local skills shortages in software engineering, cybersecurity and data roles mean many Australian employers can’t simply backfill senior positions even if they wanted to. For these firms, AI is functioning less as a replacement for scarce talent and more as an augmentation layer that lets smaller teams take on more work — a dynamic the study’s entry-level and engineering figures reflect closely.
CIOs are increasingly being asked to defend AI line items in board papers, often under pressure to show near-term cost savings. This data gives them a different argument to make: that AI spend should be measured against output, revenue and capacity gains rather than treated as a de facto cost-cutting exercise.
Governance frameworks Australian enterprises already work within, including the Essential Eight and, for regulated entities, APRA CPS 234, are built around how technology is adopted and controlled — not whether it justifies smaller teams.
Framing AI investment purely as a headcount lever sits awkwardly alongside that governance logic, and boards used to reading AI spend through a compliance lens may need to apply the same rigor to how they read its return on investment.
More must-read AI coverage
- SS&C Intralinks DealCentre AI vs. Datasite: Which platform is built for the future of dealmaking?
- SS&C Intralinks FundCentre AI vs. Juniper Square: Which platform better supports modern private markets fund managers?
- Why Data, Not Models, Determines AI Success
- The Rise of the AI-Native Factory: How Physical AI Is Transforming Manufacturing
Lessons for Australian boards
The firms in the study that saw the strongest results weren’t running perpetual pilots. They had moved AI into standard workflows across engineering, customer service, sales and finance, and they gave the gains time to compound rather than expecting an immediate return.
For Australian boards weighing sustained AI investment against near-term cost pressure, a few questions are worth asking before the next budget cycle:
- Is the organization’s AI spend still in pilot mode, or has it moved into core workflows where gains can compound?
- Are hiring decisions being made independently of AI investment, or is AI being used as a justification for cuts the business would be making anyway?
- Where skills are scarce, is AI being deployed to extend the capacity of existing teams rather than to replace roles the business can’t fill anyway?
None of this settles the broader debate about AI and job losses. The study’s authors are clear that some roles will still change or disappear as adoption spreads. But for Australian enterprises that currently treat AI budgets and headcount as opposing lines in the same spreadsheet, the data suggest that pairing sustained AI investment with a hiring freeze may be solving the wrong problem.