Concerns about generative artificial intelligence (AI) disrupting the labor market, particularly white-collar occupations long viewed as resistant to automation, have not yet materialized into widespread job losses, according to recent data analysis focused on the Australian workforce. Instead, AI is reshaping how work is performed rather than eliminating roles.

Data from Jobs and Skills Australia (JSA) assigns occupations two key metrics: an automation exposure score, indicating the proportion of tasks potentially replaceable by AI, and an augmentation score, reflecting how AI can assist workers to perform their jobs more efficiently without removing the need for human labor.

Occupations such as clerical and administrative workers, analysts, and marketing professionals typically score high on automation exposure—above 0.55 on the 0-to-1 scale—while roles involving physical tasks or interpersonal skills, like aged care workers, tradespeople, and machinery operators, tend to have scores between 0.2 and 0.3. However, these scores indicate the level of potential AI impact on tasks rather than predicting layoffs.

Tracking employment trends from February 2023 to February 2026 using Australian Bureau of Statistics figures, total employment increased by nearly 777,000 jobs, growing from approximately 13.6 million to 14.6 million workers. Of 357 occupations analyzed, most experienced growth, with no clear correlation between the degree of AI exposure and employment declines or gains.

Notably, occupations deemed highly exposed to generative AI have expanded. Software developers, with an automation score of about 0.63 and an augmentation score near 0.77, added around 37,000 jobs during this period. Similarly, advertising and marketing professionals (automation score 0.54) grew by over 21,000 jobs, while information officers (automation score 0.66) saw an increase of nearly 17,000 positions. These findings challenge expectations that AI would rapidly displace workers in these fields.

The evolving role of AI in these sectors primarily supports human workers rather than replaces them. For example, software developers increasingly use AI tools to assist with coding, enabling a shift toward validation and integration. Marketers can produce content more quickly, allowing efforts to focus on strategy and coordination.

Conversely, occupations with lower automation scores—typically between 0.15 and 0.35—have experienced declines. These roles often involve manual processes or are closely tied to economic cycles and technological shifts unrelated to AI, such as mechanization and changing production methods.

Trends in clerical and administrative jobs illustrate a more nuanced shift. Although these roles have high automation exposure scores (0.6 to 0.7), their growth slowed to 3.2 percent, below the average rate for all occupations. This slowdown impacts demand for traditional office space, as clerical roles historically drove desk density. As routine tasks become more efficient and mobile, office environments are transitioning toward spaces designed for collaboration, problem-solving, and training.

Retail employment, with automation scores between 0.45 and 0.5, showed marginal growth of 0.8 percent. The sector continues to face structural pressures from online retail expansion, changing consumer behavior, and cost concerns, with AI influencing customer engagement but not yet dominating employment trends.

Strong employment growth has occurred in occupation groups with minimal automation exposure, such as community and personal services, trades, logistics, and infrastructure roles. Aged and disabled carers, with low automation scores of about 0.2 to 0.25, increased by over 93,000 jobs, driven largely by demographic and policy factors rather than technology. Trades and machinery operators also saw steady growth between 4 and 7 percent. These trends support ongoing demand for healthcare, childcare, retail, and industrial property.

Overall, the data suggests a divergence rather than disruption: knowledge-based roles are expanding while evolving in task execution, and place-bound roles rooted in demographic needs and policy are growing strongly. The workforce is increasing by roughly 259,000 net jobs annually, but the nature of work—how, where, and why people perform their jobs—is shifting.

In summary, the Australian labor market does not yet show signs of AI-induced job losses. Instead, generative AI appears to be augmenting tasks in existing roles, leading to changes in workplace practices and physical space requirements. Commercial property markets are adapting to these shifts, with offices moving away from routine desk work toward environments fostering interaction, while sectors anchored in physical presence drive demand for healthcare, retail, and industrial facilities.