AI/Nov 16 edition/Reporting & analysis
BusinessResearchPolicy

BusinessMarkets, money & strategy

Quiggin argues NAB, Qantas and Canva job cuts reflect cost-cutting and offshoring, not AI

Economist John Quiggin argues that 2026 job cuts at NAB, Qantas and Canva mostly reflect ordinary cost-cutting, offshoring and slowing growth rather than AI. Only Atlassian clearly tied its cuts to AI, and Australian government data shows slower growth, not losses, in exposed occupations.

From the The Guardian · AI article of November 15, 2026 and cited sources: The four employers gave different reasons for their cuts. [2] [4] [5] [6]

SOURCES[2] TechCrunch – Atlassian cuts techcrunch.com[4] The New Daily – NAB to slash 2,500 jobs thenewdaily.com.au[5] ACS Information Age – NAB cuts ia.acs.org.au

AI-assisted, not human-edited. Methodology Corrections Who owns this site

KEY FACTS
What happened
The four employers gave different reasons for their cuts. [2] [4] [5] [6] [7] [8] [11]
Original source
The Guardian · AI article, independent perspective
Brief by
AT Pulse research desk AI-assisted; publisher and contact on the desk page
When
November 15, 2026 article published
Evidence
11 sources across 11 sites

Reach: not measured yet

Illustration from the Guardian: Quiggin argues NAB, Qantas and Canva job cuts reflect cost-cutting and offshoring, not AI
Image: the Guardian — Original article ↗
THE CORE IDEAS3 TAKEAWAYS
01

The four employers gave different reasons for their cuts. Atlassian cut about 1,600 roles in March to pay for AI and enterprise-sales investment. NAB's plan to cut up to 2,500 roles cited rapid technological change but did not mention AI. Canva's restructure affects 1,271 roles (22%), including role changes and relocations, and came with a growth forecast lowered from 30% to 20%. Qantas is moving 850 roles to Accenture, mostly in India, and calls the move AI-driven modernisation. That description is the company's own. [2] [4] [5] [6] [7] [8] [11]

02

Quiggin calls NAB's cuts business as usual and says Qantas's AI framing muddies what is simple offshoring. He sees Canva as the most exposed of the four, because AI makes easy-to-use design software easier to copy, and suggests AI may hit share prices before it hits jobs. [1]

03

Labour data points to slower growth rather than mass job losses. In Australia, employment in the most AI-exposed fifth of occupations grew 5.6% between Nov 2022 and Feb 2026, against 9.5% for the least-exposed fifth. US payroll data shows a widening gap for workers aged 22–25 in highly exposed jobs, driven by fewer hires rather than layoffs. [3] [9] [10]

WHY IT MATTERS

So far the evidence shows AI's effects arriving through slower hiring, especially of young workers, rather than mass redundancies.

Read the full assessment

For leaders, that means checking any AI claim in a layoff announcement against offshoring, restructuring costs and growth guidance, and watching entry-level hiring.

Executive brief

Four large Australian employers announced deep job cuts in 2026, but only one clearly said AI was the reason. Atlassian cut about 1,600 staff in March to pay for AI investment. NAB left AI out of its 2,500-role cut. Canva tied its cut of 22% of staff to slowing growth. Qantas paired offshoring 850 roles to Accenture with an "AI-ready" message. In a Guardian opinion piece, economist John Quiggin argues the cuts are mostly ordinary cost-cutting and offshoring, and that AI may hit share prices before it hits jobs. Government data from July 2026 points the same way: so far there is slower growth in exposed occupations, not broad job losses.

What changed and event timeline

‘Four years after the launch of ChatGPT, and the pattern seems to be much the same: lots of hype and excitement but with very little productivity growth.’ Illustration: BlackJack3D/Getty Images
Image: the Guardian — Original article ↗
  1. Atlassian cuts ~10%

    About 1,600 roles were cut. CEO Mike Cannon-Brookes said the aim was to self-fund AI and enterprise-sales investment, which makes this the most direct AI link of the four.

  2. Government finds no broad AI upheaval

    The Office of the Chief Economist found exposed occupations grew more slowly than others. It found no economy-wide disruption.

  3. NAB announces up to 2,500 cuts

    The bank expects 1,500 to go by the end of 2026 and a restructuring charge of about A$300m. It cited "rapid technological change" but did not mention AI.

  4. Canva restructures

    1,271 roles (22%) are affected through layoffs, role changes or relocation. The 2026 revenue-growth forecast was cut from 30% to 20%.

  5. Qantas outsources 850 roles

    The work goes to Accenture, mostly in India, with first changes due by mid-2027. Qantas framed it as modernising its systems for AI.

  6. Quiggin's rebuttal

    He calls NAB's cuts business as usual and Qantas's AI framing a way to muddy simple offshoring. He sees Canva as the most exposed of the four.

Capabilities and access

None of the four companies has disclosed which AI models or systems would take over the eliminated work.

Read the full section

None of the four companies has disclosed which AI models or systems would take over the eliminated work.

  • Qantas: says it runs about 1,000 poorly connected legacy applications. The Accenture deal is meant to modernise them, with services to be "changed through automation and AI over time" (AeroTime; Yahoo Finance).
  • Canva: recently launched Canva 2.0 and Craft, described as its own foundation model (People Matters). No version details or benchmarks are published.

Technical analysis for researchers and developers

  • Stanford "Canaries" update: uses ADP payroll data from Nov 2022 to Jun 2026. It compares workers aged 22–25 in high- versus low-AI-exposure occupations.
  • Australian study (DEWR): builds on Jobs and Skills Australia's generative-AI exposure scores, ABS Labour Force data and the Internet Vacancy Index (DEWR).
  • What this means for interpreting the cuts: both studies are exposure-based correlations, and neither can attribute individual company layoffs to AI.
Read the full section
  • Stanford "Canaries" update: uses ADP payroll data from Nov 2022 to Jun 2026. It compares workers aged 22–25 in high- versus low-AI-exposure occupations. The gap widened from 15% to 19%, driven by fewer hires rather than more layoffs. Losses cluster where AI substitutes for workers' tasks rather than complementing them (Stanford Digital Economy Lab).
  • Australian study (DEWR): builds on Jobs and Skills Australia's generative-AI exposure scores, ABS Labour Force data and the Internet Vacancy Index (DEWR).
  • What this means for interpreting the cuts: both studies are exposure-based correlations, and neither can attribute individual company layoffs to AI.

Claims and evidence

  • NAB: says the cuts are not directly about AI.
  • Qantas: "AI-driven overhaul" is the company's own description.
  • Canva: CEO Melanie Perkins did not explicitly blame AI (People Matters).
Read the full section
  • NAB: says the cuts are not directly about AI. The Finance Sector Union says executives would not confirm either way, and about 180 roles are being offshored to Vietnam and India (ACS).
  • Qantas: "AI-driven overhaul" is the company's own description. Quiggin disputes it (Guardian).
  • Canva: CEO Melanie Perkins did not explicitly blame AI (People Matters).
  • Australian labour market: employment in the most-exposed fifth of occupations grew 5.6%, against 9.5% in the least-exposed fifth (Nov 2022–Feb 2026) (DEWR).
  • Quiggin's claims: "very little productivity growth" and open-weight models running "a year behind" the leaders are his opinion. The article gives no data for either.

Context and prior work

  • Quiggin compares today to Solow's 1987 productivity paradox and the dotcom-era fibre buildout (Guardian).
  • A Yale study found no discernible disruption to the broad labour market 33 months after ChatGPT launched, with early cracks among young workers (BankInfoSecurity).
  • Atlassian followed Block, which cut about 4,000 staff in February and explicitly cited AI automation (TechCrunch).
Read the full section
  • Quiggin compares today to Solow's 1987 productivity paradox and the dotcom-era fibre buildout (Guardian).
  • A Yale study found no discernible disruption to the broad labour market 33 months after ChatGPT launched, with early cracks among young workers (BankInfoSecurity).
  • Atlassian followed Block, which cut about 4,000 staff in February and explicitly cited AI automation (TechCrunch).

Limitations, safety and contested findings

  • Conflicting accounts: Quiggin says some of the cuts were "explicitly blamed on AI." In the reviewed sources, only Atlassian clearly did so.
  • Canva's number is broader than layoffs: the 1,271 figure includes role changes and relocations (TechRepublic).
  • The research is cautious about itself: Stanford's authors say no single study is definitive and other explanations remain possible (Stanford).
Read the full section
  • Conflicting accounts: Quiggin says some of the cuts were "explicitly blamed on AI." In the reviewed sources, only Atlassian clearly did so.
  • Canva's number is broader than layoffs: the 1,271 figure includes role changes and relocations (TechRepublic).
  • The research is cautious about itself: Stanford's authors say no single study is definitive and other explanations remain possible (Stanford). The DEWR evidence is described as an "early indication" of modest slowing in some highly exposed occupations (DEWR).

Business and practitioner implications

  • Treat "AI" in a layoff announcement as a claim to check. Look for offshoring share, cost charges and growth guidance.
  • Hiring is the channel to watch.
  • SaaS pricing is under pressure where AI makes the product easy to replicate. Atlassian's organisation-wide workflow tools look harder to replace.
Read the full section
  • Treat "AI" in a layoff announcement as a claim to check. Look for offshoring share, cost charges and growth guidance. Qantas and NAB both send work to India.
  • Hiring is the channel to watch. The Stanford data shows the effect coming through fewer entry-level hires, so pipelines for junior staff are the early warning sign, not headline redundancies.
  • SaaS pricing is under pressure where AI makes the product easy to replicate. Canva's lowered growth forecast fits Quiggin's argument that easy-to-use design software is exposed. Atlassian's organisation-wide workflow tools look harder to replace.
FOLLOW THE EVIDENCE

The source trail.

Sources (11)
A LITTLE LESS NOISE. A LOT MORE CONTEXT.

Stay curious.
Follow the evidence.

Independent perspectives, the original sources, and room for the questions that don't have easy answers.

How we build the brief
Connect with us

Find us where you already read.