NextFin News - Reserve Bank of Australia Governor Michele Bullock said unemployment will need to rise to between 4.5% and 5% before inflation pressure eases, using her final public appearance before next week's rate decision to warn that the labour market remains too tight even as she flagged financial-stability risks from the AI investment boom and a crackdown on private-credit lending standards.
Speaking at a Committee for Economic Development of Australia event in Sydney on Tuesday, Bullock delivered the clearest public signal yet that the central bank is prepared to accept a softer labour market to return inflation to target. The Monetary Policy Board meets on September 28-29, and the August employment figures due Thursday from the Australian Bureau of Statistics will be the last major data point before that meeting.
The message was unambiguous. "Between 4.5 and 5 [per cent unemployment] will probably take enough heat out of the labour market that eases pressure on inflation," she said. With the unemployment rate at 4.5% in July, the governor was effectively signalling that further job-market weakening is a policy requirement, not an unwanted side effect.
The stance has already moved markets. Money-market pricing indicated roughly a 90% probability of a 25-basis-point rate increase at next week's meeting, with about even odds of a second move in November. The Commonwealth Bank of Australia lifted its two-year fixed home-loan rate by 0.48 percentage point to 6.82%, and other fixed rates rose between 0.15 and 0.3 percentage point; the other three major banks made similar moves of up to 0.2 percentage point the prior week. The S&P/ASX 200 had fallen nearly 4% in September, wiping out its year-to-date gains and giving back the record high of about 9,297 points reached in August, while the Australian dollar eased toward 0.70 US cents.
But the jobs message was only one of three threads Bullock wove together. She described artificial intelligence as "the great white hope to improve productivity" while warning that the AI boom is adding to demand ahead of any supply-side benefit — an "awkward sequencing event" for an economy already running hot. And she weighed in on the private-credit crackdown led by the Australian Securities and Investments Commission, offering a more contained view of domestic risks than the corporate regulator's stark warnings.
A Jobs Market That Must Weaken — and Why 5% Is the Line
The RBA's tolerance for higher unemployment marks a shift in emphasis. For much of 2025 and early 2026, the bank framed labour-market cooling as something to be managed gradually. Now the governor is naming a range.
The mechanism runs through wages. A tight labour market pushes up wages, which flow into business costs and then into prices. Bullock said the jobs market was "probably still a bit too tight" and putting upward pressure on wages, business costs and inflation more generally. Her concern is not just the current inflation print but what happens next: "We need to be very careful to ensure that policy is set in a way that minimises the second round and indirect effects which might perpetuate ongoing inflation."
The backdrop explains the urgency. Australia's consumer-price index rose 3.5% in the year to July, down from 3.8% in June, the statistics bureau reported on August 26. But the trimmed-mean measure of underlying inflation held at 3.6% annually and rose 0.5% in the month — a 12-month high and roughly double forecasts. The gap between a falling headline and stuck underlying inflation is the entire policy debate: the RBA's target is 2-3%, but it places most weight on the underlying measures when judging persistent pressure.
This is the classic central-bank trade-off made explicit. The RBA is choosing to lean against inflation expectations rather than protect employment in the near term. "I think monetary policy really just needs to continue to focus on making sure that we limit indirect effects and we try to keep inflation expectations anchored," Bullock said.
The bet is that a controlled rise in unemployment — to 5% — breaks the wage-price feedback before it becomes embedded. The alternative, in the RBA's framing, is worse: if expectations drift, the bank would need even higher interest rates and unemployment to bring inflation back to target, a line Bullock has used in previous speeches.
AI: Productivity Promise Today, Demand Shock First
On artificial intelligence, Bullock struck a tone that captures the central bank's dilemma. The technology should, in time, ease inflation by lifting productivity. But that benefit sits on the far side of a costly build-out phase that is stoking demand right now.
"AI is the great white hope to improve productivity. When I talk to colleagues around the world about this, I think everyone thinks that."
Then came the caveat: "But I think everyone also agrees that there are very few signs yet that AI is actually influencing the supply side of the economy. There's lots of evidence that it's influencing the demand side of the economy."
The sequencing matters. Data centres, power infrastructure and chips require capital expenditure now, while the productivity payoff arrives only once the technology diffuses through workflows. In Australia's case, that means the AI boom is adding fuel to an economy the RBA has already judged to be in excess demand. "So we've got this sort of awkward sequencing event at the moment where, in Australia at least, we are in a situation of excess demand and the AI boom is adding to that demand ahead of any potential supply impacts that it might have going forward."
The global scale of that spending is unprecedented. Venture investors poured $300 billion into about 6,000 startups globally in the first quarter of 2026, up more than 150% on both a quarterly and annual basis, according to Crunchbase data. Four of the five largest venture rounds ever recorded closed in that quarter alone — OpenAI at $122 billion, Anthropic at $30 billion, xAI at $20 billion and Waymo at $16 billion — collectively $188 billion, or 65% of global venture investment.
That concentration is why central banks are watching. "All central banks are a little bit worried about that," Bullock said. "You know, it might not be a bubble, but it might be. And if it unwinds in a disorderly manner, then that could have implications for the financial system and also for the real economies." She declined to take a position — "Some people think it's a bubble, some people don't. I don't have a particular view one way or the other, but it's a risk that I think we're watching" — but the warning landed alongside the announcement that the RBA's twice-yearly Financial Stability Review would be published the following week.
The labour-market implications of AI are more nuanced than the headline job-loss numbers suggest, Bullock said in a speech this week to the University of Western Australia's Shann Memorial Lecture. "While many experts anticipate a net increase in jobs, it is likely to be more nuanced." A survey of businesses through the RBA's liaison program found firms expect AI and robotics to "augment labour, automating repetitive tasks and redesigning the composition of roles," with headcount initially rising, then "followed by a small decline" as adoption matures. Lower-skilled roles may fall while demand for higher-skilled workers grows. "Adjustments like this can be challenging for individuals," she said, adding that displaced workers would "likely need support through these disruptions" — while noting that such support falls outside monetary policy.
Private Credit: A Regulator's Crackdown Meets a Central Bank's Calm
The third thread — private credit — exposed a divergence between Australia's corporate regulator and its central bank.
ASIC Commissioner Simone Constant, speaking at a Commercial and Asset Finance Brokers Association event in Sydney on the same day, delivered a stark assessment. The domestic private-credit sector has grown about 500% over the decade, she said, but "governance, controls and underwriting standards have not kept pace with this growth." An ASIC review of 28 funds found only four published the interest rates or ranges they charged borrowers, less than half had detailed written credit or impairment and default-management policies, and most lacked adequate separation between those approving loans and those managing them. "We challenged the sector to lift standards by 2027 — that's three months away," Constant said.
The collapse of NSW property developer Bathla has put those weaknesses under stress, with CVS Lane's First Mortgage Fund and Property Finance Fund exposed across nine loans. Constant warned that a contagion or run on funds remained "possible" if investors grew anxious about retrieving their money.
Bullock, questioned at a Senate estimates hearing on Friday, offered a more measured view. "Our assessment is that, on the whole, the non-bank lending activity we've observed is helping to, for instance, provide the financing that's important for greenfield construction, but we don't have a strong sense that there's been a systematic weakening of lending standards," she said.
The difference in tone reflects the difference in mandate and scale. ASIC is the conduct regulator looking at firm-level practices; the RBA is the financial-stability authority assessing system-wide risk. And by the RBA's own measure, Australian private credit remains small — accounting for about 11% of business lending, or roughly 2.5% of total business debt, with non-bank lenders representing about 6% of financial-system assets. That size is precisely why the RBA judges the risks "constrained," even as it watches.
The Transmission: Why These Three Issues Are One Story
The common thread is not obvious at first glance, but it is the reason Bullock addressed all three in the same week. Each issue is about the gap between financial conditions and the real economy — and about what happens when that gap closes abruptly.
A tighter labour market is the deliberate channel through which monetary policy slows demand. The AI boom is an unexpected source of additional demand that arrives before its supply-side offset. Private credit is a shadow channel of lending that could transmit stress back into the real economy if it seizes up. In each case, the RBA is trying to manage the pace of adjustment so that inflation returns to target without a disorderly break in activity or asset prices.
That is a narrower path than the one the bank walked in 2023-24. Back then, the problem was catching up with inflation after holding rates too low for too long. Now the bank is trying to fine-tune a soft landing while a global technology boom and a fast-growing non-bank credit market add uncertainty to both the demand and the financial-stability sides of the ledger.
Cyclical, Structural, or Both: The Call That Decides the Outcome
The three risks Bullock outlined are not the same kind of problem, and confusing them is the fastest way to misread the policy path. The labour-market tightening is cyclical: it is the intended, mean-reverting consequence of higher rates, and it will unwind once inflation returns to target and policy is eased. History offers a template — Australia's unemployment rate has moved through wide ranges over past cycles, and a rise from 4.5% to 5% is a modest adjustment by historical standards, not a structural break in the labour market.
Private credit sits in between. The disintermediation of lending away from banks is a structural shift in the financial system, driven by regulation and investor demand for yield. But the current stress test — Bathla, CVS Lane, ASIC's findings — is a cyclical episode: the first real downturn-like conditions the sector has faced since its rapid growth. The question is whether the structural shift survives its first cyclical test without forcing a regulatory reset.
AI is the genuinely structural force. General-purpose technologies do not revert; they reorganise the economy. The near-term demand surge is cyclical and could deflate if valuations reset. But the productivity channel, if it materialises, is permanent — and that is the second-order implication most investors are underweighting.
If AI delivers even a fraction of its promised productivity gain, it would ease inflation at the source rather than through demand destruction. That would mean the RBA's 5% unemployment threshold could prove too pessimistic: the supply-side boost would do some of the disinflationary work the central bank is currently trying to achieve by slowing the economy. Conversely, if the boom unwinds disorderly before the productivity arrives, the RBA would face the worst combination — weaker growth and still-anchored-but-fragile inflation expectations — and would have to choose between its inflation target and financial stability. That trade-off, not the next 25 basis points, is the real story.
The Counter-Case: Is the RBA Tightening Into a Slowdown That AI Will Fix?
The strongest argument against the RBA's stance is that it is over-tightening into an economy where the supply-side benefits of AI are already understated in the data. Proponents of this view — including several technology-sector economists and investors who have argued that AI-driven productivity is arriving faster than official statistics capture — contend that forcing unemployment to 5% risks a deeper-than-necessary recession. In their reading, the trimmed-mean print of 0.5% in July is a lagging indicator, and the next wave of AI-enabled automation will push prices down before the labour market needs to absorb that much pain.
There is also a financial-stability counter-argument to the RBA's calm on private credit. ASIC sees stress fractures at the firm level that the central bank's aggregate lens may miss; if losses cascade through superannuation funds and non-bank lenders, the resulting credit crunch could do the RBA's tightening work for it — but at a much higher cost to growth.
Both counter-theses have merit, but they require specific evidence to overturn the base case. The falsifying signal is clear: if the monthly trimmed-mean inflation print falls to 0.2% or below for two consecutive months while the unemployment rate climbs above 5%, the RBA will have over-tightened and will need to pivot faster than it currently signals. Until that combination appears, the bank's stated tolerance for a weaker labour market should be taken at face value.
What Comes Next: Three Horizons to Watch
In the short term, the path is already priced: a September rate rise, repriced fixed mortgages, and an equity market that has given back its 2026 gains. The August employment report on Thursday and the RBA's Financial Stability Review next week are the two immediate catalysts.
Over the medium term, the question is whether unemployment drifts toward 5% smoothly or jumps on a shock. The four major banks are split on the path — NAB anticipated a September move, CBA and Westpac now expect one with a possible second in November, and ANZ forecasts two hikes taking the cash rate to 4.85%, the highest since late 2008. ASIC's 2027 standards deadline for private credit, three months away, is the regulatory trigger to watch.
Over the long term, the structural forces dominate: whether AI converts investment into measured productivity, and whether private credit becomes a permanent, well-regulated part of the financial system or a source of recurring stress.
The base case is a controlled slowdown: unemployment drifts toward 5%, inflation expectations stay anchored, and the September rate rise is the first of perhaps two or three moves. The upside case is that AI-driven productivity arrives sooner than expected, lifting supply enough to ease inflation without as much labour-market pain. The downside case is that the AI investment boom proves to be a bubble that unwinds disorderly, or that private-credit stress spreads beyond isolated funds, forcing the RBA to choose between its inflation target and financial stability.
The central judgment: Bullock is no longer managing a gradual cooling. She is explicitly trading higher unemployment for anchored expectations, while betting that the AI boom and private-credit growth can be contained before either forces her hand.
The RBA is asking the labour market to do the heavy lifting on inflation — and hoping the AI boom and private credit don't make that job any harder.
Explore more exclusive insights at nextfin.ai.
