Women Make Up Less Than 30% of AI Job Expansion
LinkedIn data shows that women accounted for just over a quarter of new AI hires in the U.S. last year, significantly lower than the approximately 50% of women in non-AI new hires.
In AI companies across 27 countries, women hold only 13% of executive-level AI positions; among new hires for "AI leaders," women make up 20%, and for "AI directors," they account for 26%.
Women are also more concentrated in clerical, administrative, and repetitive information processing roles, which are more susceptible to the impact of generative AI. Related analysis indicates that about 9.6% of women's jobs fall into the highly AI-exposed category, nearly three times the corresponding ratio for men.
AI employment is not an overall net decrease but shows a divergence of "high-paying technical job expansion and pressure on some entry-level knowledge work." Data shows that the employment rate of young people aged 22 to 25 in high AI-exposure occupations is about 16% lower than their peers in low-exposure jobs.
AI-related positions are typically concentrated in model development, data engineering, MLOps, product, computing power, cybersecurity, and infrastructure, where women have lower representation in educational pathways, technical recruitment channels, and senior management positions, creating a dual gap of "being more easily replaceable and having less access to new positions."
From a market mechanism perspective, companies prioritize AI budgets for engineering, data, cloud computing, and automation roles that can directly enhance productivity, while compressing administrative and basic content processing jobs that can be replaced by tools. The beneficiaries are those with technical training, data access, and decision-making power; the ones under pressure are employees concentrated in low-bargaining, process-oriented office roles. If recruitment, training, promotion, and AI tool allocation do not change, the returns on AI investment will flow more to the existing pool of technical talent, while the exposure and opportunity gaps faced by women may widen simultaneously.
Source: Public Information
ABAB AI Insight
Women are not inherently at a disadvantage in the technological revolution; the key lies in their position within the career chain. Personal computers, the internet, and mobile platforms have all created new jobs, but high-growth positions are typically concentrated first in engineering, product, capital, and management. If women are more concentrated in administrative, coordination, and support roles, they are likely to bear the brunt of process automation while sharing less in the equity, salary, and promotion benefits of new positions. The difference with AI is that generative tools have for the first time entered text, customer service, documentation, and coordination tasks on a large scale, directly impacting traditional white-collar roles where women have a higher proportion.
The capital pathway is dividing "knowledge work" into two categories: those who own models, data, computing power, and workflow design rights gain higher output and bargaining power; those executing standardized input, organization, scheduling, and basic content generation face automation replacement or wage pressure. Corporate AI spending flows into cloud services, GPUs, data platforms, software engineering, and AI product teams, where female representation is low, making it harder to balance the new salaries, management rights, and equity incentives that come from capital expenditure.
This can be contrasted with the digital transformation in finance, law, and consulting: Excel, ERP, and automation first eliminate or compress a large number of repetitive backend tasks but enhance the value of roles that can interpret data, configure systems, and lead client relationships. AI will not directly eliminate all administrative or content work; rather, it will raise job requirements from "completing processes" to "defining processes, reviewing model outputs, handling exceptions, and taking responsibility for results." Whether women can enter the AI dividend layer depends on whether they gain technical training, management promotion, tool usage rights, and project responsibility, not just whether they learn prompt engineering.
Essentially, this is a combination of technological replacement and capital concentration: technological replacement compresses standardized positions, while capital concentration determines who gains new positions and benefits. The mechanism at play is that training data, computing power, model deployment, and high-paying technical positions are already concentrated in a few companies and professional groups; if education, recruitment, and internal promotion continue to rely on old networks, AI will amplify existing gender divisions. The solution is not to place women solely in low-value AI support roles but to increase their participation in model development, data governance, product decision-making, corporate procurement, and board-level discussions.
ABAB News · Law of Cognition
Jobs replaced by AI are often not those that create AI.
Technological dividends are not automatically shared; access determines distribution.
Knowing how to use tools is just efficiency; defining tools grants power.