26% vs 50%: The AI Jobs Gender Gap in US Hiring Today

The AI jobs gender gap in the United States now has a hard number attached to it: 26%. That's the share of women among people hired into roles requiring artificial intelligence skills, according to research published this month by LinkedIn's Economic Graph Research Institute. In every other category of work, women account for half of new hires. The gap opens precisely where the money is.

And there is a lot of money there. A typical AI job listing advertises around $177,000 a year, against roughly $80,000 for a listing with no AI component — a spread of about $97,000 per posting. The number of AI job ads has doubled since 2023. So the fastest-growing, best-paid corner of the American labor market is being staffed at roughly three men for every woman.

That matters beyond fairness, because the same workers are exposed from the other direction. Brookings Institution researchers estimate 6.1 million Americans hold jobs with high AI exposure and little capacity to absorb a shock — mostly clerical and administrative work, and about 86% of the people doing it are women. One door is closing faster than the other one is opening.

Where the 26% Number Comes From

LinkedIn built the figures from its own hiring and posting data across 27 countries between 2023 and 2026, released as two reports: The Triple Penalty and The AI Talent Divide. These count hires, not existing headcount, which makes them a leading indicator — they show who is walking through the door right now. The intake also skews young: Gen Z accounts for more than two-thirds of AI engineer hires.

The Triple Penalty, Broken Down

The phrase comes from LinkedIn's own title, and it describes three disadvantages stacking on top of one another. Each is modest alone. Together they explain why the executive floor of an AI company looks so different from the executive floor of an ordinary one.

Penalty One: The Company You Work For

Women are about 5 percentage points less represented at AI companies than at non-AI companies. It's the smallest of the three and easy to overlook, but it means the gap begins before anyone applies for a specific job — it's baked into which employers are growing. That pattern showed up in 24 of the 27 markets studied.

Penalty Two: The Role Itself

Inside any given employer, AI roles run about 10 percentage points behind non-AI roles on women's representation. This is the layer most people picture when they hear "tech gender gap," and the one retraining programs aim at. It's also where the salary premium lives, which is why this layer costs the most in lost income.

Penalty Three: The Jump to the C-Suite

The last step is the steepest. Women are roughly 15 percentage points less represented in executive AI roles than non-executive ones, and hold just 13% of C-suite seats at AI companies — a gap that appeared in all 27 countries. Hiring titles tell it plainly: 20% of Head of AI hires, 26% of Director of AI hires, and 18% of member of technical staff hires, the catch-all title leading labs use for core research and engineering work.

Why the AI Jobs Gender Gap Is Hard to Close

Credentials do much of the gatekeeping. About 91% of AI roles are held by workers with at least a bachelor's degree. Among Head of AI hires it's 98%, roughly 70% hold a graduate degree, and around 20% have a doctorate. When a job screens that hard on advanced degrees in a narrow band of fields, it inherits whatever imbalance those fields carried twenty years ago.

Those fields shrank. Women earned 37% of US computer science degrees in 1984 and about 18% by 2014. Their share of professional computing jobs peaked near 36% around 1991 and had slipped to roughly 26% by 2019. The AI jobs gender gap isn't a fresh failure so much as an old one, amplified by a hiring wave moving faster than any pipeline can answer.

The One AI Job Where Women Are the Majority

There is an exception, and it's revealing. Data annotation — labeling the images, text and audio that models train on — is the most gender-balanced AI occupation in the study, with women taking more than half of hires. It's also among the lowest paid. Those jobs demand less formal education and are more often hourly, project-based or remote. Women are entering AI work; they're entering the part that pays least.

What to Do If You Want One of These Jobs

LinkedIn's proposed fix, laid out by company spokesperson Sarah Steinberg, is skills-based hiring — screening for what a candidate can do rather than which degree they hold. That's a reasonable ask of employers and a slow one. Meanwhile, the individual move is to shorten the distance between the job you have and the posting you want.

• Read ten real AI listings in your field and write down every tool they name. That list is your syllabus.
• Attach AI work to your current title — automating a report, building an internal assistant — so your résumé shows shipped work, not certificates.
• Check adjacent titles. AI product, AI operations and model evaluation roles often want domain judgment more than a doctorate.
• If you work in administrative or clerical support, treat this as urgent rather than optional.

The figure to watch is whether 26% moves. LinkedIn refreshes this data, and the next release will show whether skills-based hiring is a real shift or a talking point. If your role sits anywhere AI is likely to touch, don't wait for that answer — these seats are being filled now, and the people getting them are the ones who can already point at something they built.