The global capitalist system is leveraging artificial intelligence to reduce demand for entry-level labor, with UK job postings data showing a sharp contraction in available positions. Adzuna, a job listing tracker, reported that entry-level vacancies fell by 31.9 percent between November 2022 and mid-2025. This decline in demand for new workers signals a systemic shift where technology is deployed to suppress wages and increase surplus extraction rather than to expand employment.
Over the same period, the share of entry-level jobs in the overall labor market decreased from 29 percent to 25 percent. This reduction indicates a structural change in the composition of the workforce, where the initial effects of AI are manifesting as weaker demand for new hires rather than immediate mass layoffs. In the United States, technology and finance firms are slowing the hiring of entry-level analysts, programmers, and support staff. This slowdown is directly linked to AI tools augmenting tasks such as research, drafting, and coding, allowing capital to achieve similar output with fewer human laborers.
Who Pays the Cost
The burden of this shift falls disproportionately on younger workers. In Australia, employment growth for junior white-collar occupations lags behind broader labor-market growth, a pattern that predates AI but is now being accelerated. Deloitte Access Economics analysis in Australia found that workers aged 15–24 made up 22.9 percent of employment in AI-disrupted roles in November 2025, compared with 15.2 percent of overall employment. This overrepresentation indicates that the capitalist drive for efficiency through AI is intensifying precarity for those entering the workforce.
The article notes that AI-driven workforce change may not yet be a major factor behind rising unemployment, which it attributes to higher interest rates and slower government spending growth. These state policies, separate from AI adoption, also contribute to the economic insecurity faced by the working class. US Bureau of Labor Statistics data released in February 2026 showed that only 181,000 jobs were added to the US economy over the course of 2025, a significant drop from 2.0 million in 2024. This modest rate of job growth occurred even as real GDP growth remained robust, illustrating the widening gap between capital accumulation and the provision of stable employment for workers.
Capital's Strategy and the State's Role
The analysis by Deloitte Access Economics outlines that AI’s largest employment impacts are likely to be in roles defined by routine automatable tasks that do not always require the application of human soft skills, including a range of clerical and information processing occupations. Employment growth in these occupations in Australia has increased at roughly half the pace of total employment growth over the past decade, reflecting both recent AI developments and earlier waves of automation. The article frames the "critical next step" for organizations as the shift from individual AI experimentation to "enterprise-wide integration," a clear directive for capital to maximize its returns by embedding AI into every facet of production and administration.
David Rumbens, a Partner at Deloitte Access Economics, advises CFOs and workforce planners to "not be complacent" and to "look at the evidence and cut through the noise." This perspective centers on managing "workforce change" from the vantage point of corporate leadership, focusing on "operational needs and sector-specific developments" rather than the structural displacement and wage suppression experienced by the working class. The forecast for employment growth to diverge across occupation groups further highlights this restructuring, with community and personal service workers, trades workers, and many professionals expected to see robust demand, while AI-exposed professions, managers, clerical and administrative staff, and sales workers may see softer demand. This divergence underscores the ongoing re-segmentation of the labor market to serve the evolving needs of capital.