El Siglo Futuro - AI Job Losses at NAB, Qantas, Canva and Atlassian Signal Market Shift, Not Workforce Collapse

The recent wave of job cuts at major Australian corporations, including National Australia Bank (NAB), Qantas, Canva, and Atlassian, has reignited fears that artificial intelligence is triggering an immediate collapse in the labor market. However, a closer examination of the evidence suggests that these reductions are largely consistent with historical business practices rather than a new era of automated unemployment. While the hype surrounding AI remains intense, its primary economic impact thus far mirrors the dot-com boom of the 1990s: a surge in stock market valuations and infrastructure investment, rather than widespread productivity gains or mass job displacement.

Four years after the launch of ChatGPT, the pattern of technological adoption follows a familiar trajectory. Nobel Prize-winning economist Robert Solow observed in 1987 that while computers were visible everywhere, they had not yet appeared in productivity statistics. Real productivity gains eventually emerged in the 1990s following the advent of the internet, but these increases were short-lived. Today, despite the excitement surrounding generative AI, significant productivity growth remains elusive. The main economic effects are instead seen in a stock market boom and massive capital expenditure on data centers, betting on future demand that has not yet materialized.

The announcements from NAB, Qantas, Canva, and Atlassian have raised questions about whether the "job apocalypse" is finally here. Yet, the specific circumstances at each firm tell a different story. NAB has explicitly denied that AI played any role in its recent workforce reductions. This denial is credible given the bank’s long-standing operational model. Since 2010, NAB has announced mass layoffs in most years, sometimes more than once. Despite these periodic cuts, total employment numbers have remained relatively stable as employees quit or retire. The bank’s strategy of hiring when needed and reducing staff when profitable is a standard feature of capitalism, marking a departure from the era when large employers expected to provide long-term loyalty to their workers.

Qantas has invoked AI to explain its decision to outsource 850 jobs to India, a move that critics argue is designed to obscure the reality of offshoring. This strategy aligns with the airline’s longstanding practice of prioritizing cost reductions over Australian labor, often justified by its status as the national flag carrier. The claim that these roles will eventually be replaced by AI complicates the narrative but ultimately weakens Qantas’s position. If automation were the primary driver, there would be little logical reason to incur the costs and complexities of establishing a new overseas operation ahead of time. This situation invites scrutiny from Australian political leaders who benefit from the airline’s executive hospitality.

Canva appears more vulnerable to AI disruption due to its business model. The company operates on a "software as a service" basis, charging subscriptions for accessible graphic design tools. While this model has been highly successful for users seeking simplicity over complex professional packages, generative AI now allows individuals to create posters and other designs in seconds. Although the output may be considered low-quality or "AI slop," many consumers are willing to accept it for basic needs, potentially eroding Canva’s value proposition.

Atlassian faces challenges as well, though its software offerings are harder to replicate through AI. The company’s applications focus on project management and organization-wide data coordination. Piecemeal use of AI by individual workers is unlikely to replace these comprehensive functions. While it is possible that AI could enable the reverse engineering of such software packages, this remains uncertain.

Finally, the hyperscalers driving the AI revolution, including OpenAI, Anthropic, Meta, and Alphabet, are not immune to market shifts. Cheap "open-weight" models from competitors like DeepSeek in China and Mistral in France are closing the technological gap rapidly while investing a fraction of the capital required by leaders. If large language models become as ubiquitous and unprofitable as email, the economic consequences could be severe for investors. In such a scenario, an apocalypse may indeed occur, but it will likely manifest in the stock market rather than the jobs market.

John Quiggin is a professor at the University of Queensland’s school of economics.

F.González--ESF