Key Highlights

  • The share of firms with no AI adoption fell from about one-third in 2025 to almost zero in 2026.
  • Slower hiring, rather than mass layoffs, is the main pattern in the liaison data.
  • Sarah Hunter's view was direct: exposure to AI does not necessarily mean Job loss.
  • The data are a snapshot of what firms say has happened so far, not a forecast.

Most businesses in the Reserve Bank's liaison program say artificial intelligence has not changed their headcount, and where it has, slower hiring is more common than job cuts.

Ask what Artificial Intelligence (AI) will do to jobs and the answers range from utopian to apocalyptic. The Reserve Bank of Australia has taken a more grounded approach: it asked businesses directly.

In a speech titled "What Happens to the Economy as 'Intelligence' Gets Cheaper?", delivered at the UBS Australasia Conference in Sydney on 9 November 2026, Assistant Governor (Economic) Sarah Hunter set out what the RBA's liaison program has heard from firms. The headline finding is undramatic, and that is precisely the point.

What firms reported on headcount

According to Hunter, when the RBA's liaison contacts were asked how AI had affected their staffing:

  • Around three-quarters reported no change to headcount;
  • About 12 per cent reported slower hiring or not filling roles;
  • About 8 per cent reported reducing headcount;
  • Around 6 per cent reported increasing headcount.

Put another way, roughly one in four firms reported some change, and among those the most common response was to hire more slowly or leave positions vacant rather than to let existing staff go. A small minority said AI had led them to add workers.

That pattern matters because it shapes how any labour market adjustment might unfold. Fewer new hires tends to show up gradually, through fewer openings and longer job searches, rather than through sudden waves of redundancies.

Adoption is spreading quickly

The muted headcount response does not reflect a lack of uptake. Hunter's liaison data show adoption has broadened sharply. The share of firms with no AI adoption fell from about one-third in 2025 to almost zero in 2026. Firms with limited adoption rose from around 40 per cent to around 60 per cent, moderate adoption from around 20 per cent to 30 per cent, and deep or transformative adoption from 8 per cent to 10 per cent.

So nearly every firm in liaison is now using AI in some form, but most are still at the limited end. That helps explain why the effects on jobs have so far been modest. Firms' reported productivity gains tell a similar story: around 35 per cent reported no gain, around 30 per cent minor gains, 23 per cent moderate gains and around 11 per cent significant gains.

Exposure is not the same as replacement

A common worry is that large parts of the workforce are exposed to AI and therefore at risk. Hunter drew a distinction using estimates from Jobs and Skills Australia. About 4 per cent of workers are highly exposed to automation from AI, meaning AI could perform much of their work. A much larger share, 31 per cent, are highly exposed to AI augmenting their work, meaning the technology is more likely to assist than replace them.

Her view was direct: "So, exposure to AI does not necessarily mean job loss."

She framed the likely outcome as a mix of effects: "AI will automate some tasks, complement workers in others and create new roles as the economy adapts." That is consistent with the liaison headcount figures, where a small group of firms is cutting, a slightly larger group is hiring more slowly, and a few are expanding.

Hunter also pointed to history. Transformative technologies, she said, have a pattern: "the largest gains from transformative technologies often take longer to emerge than initially expected." And she acknowledged that not every use of AI adds value: "The term 'AI slop' has entered our vernacular to capture unproductive uses."

Fitting AI into the jobs picture

The RBA's broader labour market view provides useful context. The November 2026 Statement on Monetary Policy reports that "The Unemployment rate increased to 4.6 per cent in the September quarter." The bank attributes that rise largely to higher-than-expected labour supply, noting that "employment growth held up in the September quarter as expected".

Nothing in the RBA's published material attributes the recent rise in unemployment to AI. The forecast path, with unemployment rising to a bit over 5 per cent by end-2028, is explained in terms of slower economic growth and restrictive financial conditions following the Monetary Policy Board's decision on 3 November to lift the cash rate target to 4.85 per cent.

On inflation, Hunter's view was that AI is unlikely to be a quick fix: "Consistent with previous general-purpose technologies, AI's long-run effect on Inflation is likely to be broadly neutral." That matters for the Board, which is focused on bringing down headline inflation of 3.9 per cent and trimmed mean inflation of 3.7 per cent over the year to the September quarter.

There is also a financial dimension. Hunter noted that "around 5 per cent of Australian households' financial Wealth is invested in the Equity of AI-related companies, both directly and through superannuation." US AI-related company share prices have "roughly quadrupled since the end of 2022", and she warned that if AI returns disappoint, such Assets "will be exposed to sharp repricing and possibly disorderly adjustments". Those risks sit alongside, rather than within, the staffing story.

The liaison figures are a snapshot of what firms say has happened so far, not a forecast. If more firms move to deeper use, the balance between slower hiring, job cuts and new roles could shift.

The bottom line

Businesses told the RBA that AI has, so far, mostly left their headcount unchanged. Around three-quarters reported no change, about 12 per cent were hiring more slowly or not filling roles, about 8 per cent had reduced staff and around 6 per cent had added workers. Adoption has spread to almost every firm in liaison, yet measurable productivity gains remain limited and uneven. Sarah Hunter's message was that exposure to AI does not necessarily mean job loss, and that the biggest effects may take longer to arrive than expected.