Key Highlights

• The 30% figure is a market-implied expectation, not a measured productivity result.

• Hunter cited a US study implying markets expect a roughly 30% permanent productivity gain for software engineers.

• US AI-related share prices have roughly quadrupled since the end of 2022.

• The RBA's liaison data show widespread but mostly limited AI use.

Financial markets appear to be pricing in a large, lasting lift in software engineering productivity from artificial intelligence, but the Reserve Bank's evidence shows that is an expectation, not yet a measured result.

The question mark in the headline matters. When Reserve Bank of Australia Assistant Governor (Economic) Sarah Hunter spoke at the UBS Australasia Conference in Sydney on 9 November 2026, she cited a US study implying that financial markets expect AI to permanently raise software engineering productivity by "around 30 per cent."

That is a striking number. But it is a market-implied expectation, drawn from how investors are pricing assets, not a measurement of what software engineers are actually producing. The gap between what markets are betting on and what the data currently show is the central theme of Hunter's speech, titled "What Happens to the Economy as 'Intelligence' Gets Cheaper?"

Reading a market-implied number

A market-implied expectation works backwards. Rather than observing output directly, it asks what investors must believe for asset prices to make sense. If share prices of AI-related companies have risen firmly, one interpretation is that markets expect AI to deliver substantial and durable gains for the businesses and workers using it.

The 30 per cent figure for software engineering is an expression of that belief. It tells us something about sentiment and the size of the bet. It does not tell us that productivity in software engineering has already risen by that amount, and it should not be read that way.

The scale of the bet is visible elsewhere in Hunter's speech. She noted that US AI-related company share prices have "roughly quadrupled since the end of 2022." In Australia, data centre Investment has risen "10-fold (from a very low base) over the past three years."

What the measured evidence shows

Set against those expectations, the evidence gathered by the RBA is more modest. Hunter was direct: "So far, there is limited evidence that AI has materially lifted aggregate productivity."

The RBA's liaison program asked firms what productivity gains they had experienced from AI. 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.

Adoption, meanwhile, has broadened quickly. The share of liaison firms with no AI adoption fell from about one-third in 2025 to almost zero in 2026. 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. Most firms, then, are using AI, but mostly at the shallow end.

Forecasts for the decade ahead are widely dispersed.

She also noted that not all use is productive: "The term 'AI slop' has entered our vernacular to capture unproductive uses."

Why software engineering, and what about other work?

The software engineering estimate is about augmentation: making existing work more productive. Hunter's broader framing of the labour market suggests that is the more common channel than outright replacement.

Citing Jobs and Skills Australia, she said about 4 per cent of workers are highly exposed to automation from AI, while 31 per cent are highly exposed to AI augmenting their work. Her summary: "So, exposure to AI does not necessarily mean Job loss." And: "AI will automate some tasks, complement workers in others and create new roles as the economy adapts."

Liaison data on headcount point the same way. Around three-quarters of firms reported no change in staffing due to AI, about 12 per cent reported slower hiring or not filling roles, about 8 per cent were reducing headcount and around 6 per cent were increasing it.

The risk if the bet disappoints

Markets pricing in a 30 per cent permanent productivity lift for one occupation are making a forward-looking call. If returns fall short, the adjustment could be abrupt. Hunter warned that in that case Assets "will be exposed to sharp repricing and possibly disorderly adjustments".

Australians have a stake in that outcome. Hunter said "around 5 per cent of Australian households' financial Wealth is invested in the Equity of AI-related companies, both directly and through superannuation."

The RBA's October 2026 Financial Stability Review reached a broadly reassuring overall judgement while flagging external threats: "Our latest assessment is that the Australian financial system remains resilient, but there's no room for complacency." It also observed: "The principal risks are increasingly coming from abroad."

History suggests patience may be needed even if the optimism proves justified. Hunter noted that "the largest gains from transformative technologies often take longer to emerge than initially expected." A market bet can be right about the size of a gain and still be early on its timing.

What it means for Inflation and rates

For the Monetary Policy Board, AI is not a short-term answer to inflation. Hunter said: "Consistent with previous general-purpose technologies, AI's long-run effect on inflation is likely to be broadly neutral."

The Board's immediate concern remains current price pressures. Headline inflation was 3.9 per cent over the year to the September quarter and trimmed mean, or underlying, inflation was 3.7 per cent; the monthly CPI indicator showed headline inflation of 4.3 per cent for the September month. On 3 November the Board lifted the cash rate target by 25 basis points to 4.85 per cent.

The bottom line

The 30 per cent figure in the headline is real, but it is a market bet rather than measured evidence: Hunter cited a US study implying markets expect AI to permanently raise software engineering productivity by around 30 per cent. The RBA's own liaison data show widespread but mostly limited adoption, uneven reported gains and little sign so far of a material lift in aggregate productivity. With US AI-related share prices roughly quadrupling since the end of 2022 and about 5 per cent of Australian households' financial wealth tied to AI-related equities, the RBA's message is that expectations are running ahead of the evidence, and that assets priced for rapid gains could reprice sharply if returns disappoint.