Mentioning AI Is Not Doing AI

Everyone says AI is transforming the nonprofit sector. I went to the IRS filings to see how big the AI nonprofit field actually is.
289 U.S. nonprofits mention AI in their mission or program text. The programs mentioning AI carried $12.2 billion in expenses in 2023.
Both numbers are misleading.
Almost all of that spending comes from hospitals and universities citing an AI tool inside much larger programs. The dollars belong to cancer research, computing, and other fields, not to an AI nonprofit sector.
Mentioning AI is not doing AI.
The organizations actually dedicated to AI work put roughly $400 million a year into operations. Half of that is a single organization.
The gap between the $12.2 billion attached to programs that mention AI and the roughly $400 million spent by dedicated AI organizations is about 30x. That gap is exactly where mission-statement screening misleads funders, advisors, and journalists.
A note on the data: nonprofits file 990s about a year after their fiscal year closes, so tax years 2023 and 2024 are the most recent reasonably complete filing years, and the dedicated-field figures are floors. The AI nonprofit field in 2026 is certainly larger than the filings show. It is not 30x larger.
Later this week: how fast this field is actually growing, and what the funding trajectory looks like.
The full SciRise report on the AI nonprofit landscape comes out next week, covering who the real field is, who funds it, and the intersections where almost nobody is working yet.
Source: SciRise Mission-Driven Capital Intelligence Dataset. SciRise is an independent venture and is not affiliated with, sponsored by, or endorsed by Duke University.
