92% of Surveyed US Nonprofits Use AI. Nearly Half Have No AI Policy.

A 2026 benchmark report from Virtuous and Fundraising.AI surveyed 346 nonprofits in late 2025. It found about 92% using AI tools in some capacity, and just 7% reporting major improvements in their organisational capability. Roughly 47% had no AI policy at all (
,
).

Treat these as an indication. The sample is small, it was run by fundraising-software vendors, and it is not a representative survey of US nonprofits. Other surveys give different policy figures because they define "policy" differently.

Why a board should care anyway

A nonprofit board's fiduciary duties, generally described as care, loyalty and obedience, do not stop at the edge of new technology. The gap usually sits with the board, not IT. Not because boards resist AI governance, but because nobody has done the diagnosis: where is AI already in use, who is accountable, and what does the policy say?

What about the NIST framework?

The NIST AI Risk Management Framework is a useful reference. It is voluntary guidance, and I could find no rule that makes it a condition of federal grants. If you receive federal funding, check the terms of each award and the funding agency's current AI guidance.

What a board can do this month

• Know, concretely, where AI already touches the organisation.
• Before AI tools touch decisions about the people you serve, ask for bias and fairness checks, especially where a blind spot would fall hardest on marginalised communities.
• Align technology choices with the organisation's values, not efficiency for its own sake.

Take the Free AI Governance Check →

Updated 16 October 2026 after re-checking every source.