Women are 26% of US AI hires and 13% of AI C-suites: the buildout is picking its bench now

A hiring share is not a headcount. The 26% LinkedIn reports covers people moved into AI roles during 2025, not the people already sitting in them ^[2], which makes it a description of decisions being signed off this year rather than a backlog someone else created. Teams converge on their intake mix. A group staffed at 26% does not drift toward the 50% its own non-AI hiring produces; it comes to look like its requisitions ^[10].

The seniority figures do not stack cleanly on top of the hiring figure, and that is worth saying rather than smoothing over: 26% is US hires, while the head-of-AI, member-of-technical-staff and C-suite shares are global, and the last is specific to AI companies ^[2][3]. Even allowing for that, the C-suite share is half the US intake share ^[9], so nothing in the data suggests the gap narrows as titles get bigger.

The review systems are where this stops being a recruiting question. Gusto tags workers with one of five archetypes each quarter, and Travelport built a dashboard letting managers see which engineers are skipping certain AI features, according to Business Insider ^[5]. Rating people on AI usage rewards proximity to AI work, and proximity is handed out by the same staffing decisions that produced the 26% ^[1][2]. A firm can hold a 26% intake and a fluency-weighted rating scale at the same time without anyone connecting the two documents.

What prices the error is Bridgewater's internal estimate that 18% of US jobs could be displaced inside five years ^[11], set against Goldman Sachs finding the hiring slowdown already visible in specific developed-economy sectors ^[4]. In that market the AI requisition is the mobility route out, and the intake number says who is being routed. Bridgewater's own proposed remedy, an early tax on AI tokens, is contested by some experts ^[11], which is another way of saying no policy backstop arrives on the timeline of a two-year hiring plan.

The outside pool is not about to loosen the constraint either. Pew finds a majority of young adults now more concerned than excited about AI, with under-30s showing the sharpest rise in concern over the past two years while concern fell among 50-to-64s ^[6]. Randi Weingarten, whose AFT wants student-facing AI tools kept away from elementary students while calling AI fluency critical for students entering the workforce ^[7], is describing a supply-side correction that pays out in a decade, not in this budget cycle.

The default is doing the work here. A leader filling AI roles from the networks that already produce AI candidates arrives at 26% without making a decision that looks like one, and inherits a 24-point gap against the company's own hiring baseline ^[8]. The measurable version of the problem is available now, per hundred hires, before the displacement estimates are tested.