The AI Funding Gender Gap: What the Grant Data Reveals

Women make up 22% of the global AI workforce, hold fewer than 20% of AI faculty positions, and authored just 13.8% of AI research papers in the last comprehensive count. Those numbers have barely moved in a decade. But the funding pipeline tells an even sharper story: at the National Institutes of Health, first-time women principal investigators receive grants roughly 24% smaller than their male counterparts, a gap wide enough to determine whether a lab survives its first five years.

Browse our AI Grants page for current opportunities in artificial intelligence research funding.

The disparity matters beyond fairness arguments. Research teams with diverse leadership produce higher-impact publications and are more likely to address health conditions affecting underserved populations. When AI funding concentrates among a narrow demographic, the technology it produces reflects those blind spots at scale.

The NIH Numbers: Smaller First Grants, Fewer Renewals

A study published in eLife examining NIH Research Project Grant data through fiscal year 2020 found that women's share of funded PI positions rose from 22% to 33% between 1998 and 2019. That sounds like progress until you examine the other side of the ledger: success rates for women applicants dropped from 30% to 21% during the same period, meaning more women are applying but a smaller fraction are getting funded.

The award size gap is especially pronounced for early-career researchers. An analysis of 53,903 first-time NIH grants found that male PIs received a median of $165,721 compared to $126,615 for women — a difference of roughly $39,000. At Big Ten universities, the gap was even wider, with men receiving a median of $148,076 versus $66,365 for women. That $40,000 to $80,000 shortfall does not just slow research. It affects hiring decisions, equipment purchases, and whether a junior investigator can generate the preliminary data needed to compete for an R01 renewal.

In fields converging with AI — biomedical imaging, computational genomics, clinical decision support — these disparities compound. The NIH's Bridge to Artificial Intelligence (Bridge2AI) program has committed to building diverse teams, but the broader NIH R01 ecosystem, which funds the majority of investigator-initiated AI-adjacent research at $250,000 to $300,000 per year for up to five years, still reflects structural patterns that disadvantage women at the starting line.

NSF's AI Portfolio: Progress in Programs, Gaps in Professorships

The National Science Foundation tells a different but related story. NSF's CISE directorate — the primary home for AI, machine learning, and intelligent systems research — has historically struggled with gender representation among its funded PIs, mirroring the composition of computer science departments where men hold 80% of AI faculty positions.

The NSF CAREER award, the foundation's most prestigious early-career recognition, explicitly encourages applications from women and underrepresented minorities. Yet an investigation by Science found that women comprised as few as 10% of top performers in some recent years, reaching 40% only in peak years — a volatility that suggests the pipeline remains fragile rather than structurally reformed. The next CAREER deadline is July 22, 2026, with awards typically running $500,000 to $600,000 over five years.

NSF's Future CoRe program (NSF 25-543), which replaced the former CISE Core Programs solicitation, funds 400 to 600 awards per cycle at up to $1 million over four years. It is the highest-volume pathway for AI researchers, with target dates in February and September annually. But volume alone does not solve the faculty pipeline problem: if women hold only 16% of tenure-track computer science positions, the eligible applicant pool is constrained before any proposal is ever written.

The ADVANCE Dismantling and Its Fallout

For 24 years, NSF's ADVANCE program funded institutional reforms designed to address systemic barriers for women in STEM. Clark Atlanta University used ADVANCE funding to restructure hiring practices. Dozens of R1 institutions overhauled tenure review processes based on ADVANCE-supported research. The program had shifted from supporting individual women to transforming the structures that disadvantaged them — a model that peer-reviewed evaluations consistently found effective.

In spring 2025, NSF terminated the ADVANCE program and canceled over 1,000 grants across its equity-focused portfolio. The Division of Equity for Excellence in STEM, which administered ADVANCE, was shuttered entirely. An analysis by Science found that these cuts fell disproportionately on researchers from underrepresented groups. A federal court later ordered partial reinstatement of some awards, but the program's institutional infrastructure — the staff, the review panels, the accumulated expertise — is gone.

For women AI researchers, the timing is particularly damaging. The field is growing faster than almost any other discipline in STEM, and the structural interventions that might have corrected its gender imbalance during this growth period have been defunded at the moment they mattered most.

What Individual Applicants Can Do Now

The data is discouraging, but the funding landscape is not monolithic. Several active programs specifically address the pipeline.

NIH Administrative Supplements for AI Workforce Development — The Office of Data Science Strategy funds supplements that add AI/ML training positions to existing NIH grants. These are not standalone awards; they attach to active R01s, U01s, and similar mechanisms, making them accessible to PIs who already hold NIH funding and want to train women and underrepresented researchers in AI methods.

NSF ExpandAI (NSF 23-506) — While primarily designed for minority-serving institutions, ExpandAI funds AI capacity building through partnerships with existing NSF AI Institutes. Track 2 awards create pathways for researchers at HBCUs, HSIs, and other MSIs to access resources and collaborations that the broader CISE ecosystem does not reliably provide.

DARPA Young Faculty Award — DARPA's YFA program targets early-career researchers without prior DARPA funding. It does not restrict by gender, but its focus on untenured faculty at the assistant professor level means it reaches the career stage where the gender gap in AI is widest. Awards are typically $500,000 over two years with potential for a $200,000 Director's Fellowship extension.

The Structural Problem Beneath the Numbers

Stanford's AI Index has described the needle on AI diversity as "unmoving." Women earn roughly 21% of computer science bachelor's degrees — down from 37% in 1984. They account for less than 19% of AI and computer science PhD graduates in North America over the past decade. At the faculty level, representation drops further still.

Grant agencies cannot fix a pipeline that higher education has not built. But they can stop penalizing the women who do make it through. Equalizing first-award sizes, blinding review stages where gender correlates with scoring disparities, and funding institutional reform programs rather than cutting them are interventions supported by the evidence. The AI research community is building systems that will shape every sector of the economy — and right now, fewer than one in five of the people building those systems are women.

For researchers navigating this landscape, Granted can help surface the programs, deadlines, and eligibility requirements that match your profile, so the search itself does not become another barrier.

Sources:

• eLife: Inequalities in NIH Research Project Grant Funding
• NIH Bridge to Artificial Intelligence (Bridge2AI)
• NSF CISE Directorate
• NSF CAREER Program
• NSF Future CoRe (NSF 25-543)
• NSF ADVANCE Program
• Science: NSF Grant Cuts Fall Heaviest on Underrepresented Groups
• NSF ExpandAI (NSF 23-506)
• DARPA Young Faculty Award
• Stanford HAI: AI Index Diversity Report
• NIH AI Workforce Development Supplements