What AI equity for nonprofits really looks like

Nonprofits are talking about AI more than ever—but practicing equity with it less. Of the 850 nonprofits in our AI Equity Project 2025 survey, 80% reported familiarity with AI and 58% reported familiarity with data equity, up from 53% and 50% in 2024. The survey also found 64% were familiar with AI bias, up from 44% in 2024, while more than half of nonprofits feared AI could harm marginalized communities.  

Yet, only 36% of respondents were implementing equity practices, down from 46% in 2024. This is not a sign of failure—it’s a sign of friction. Awareness doesn’t automatically translate into readiness. The question is not just “Are we learning about AI?” but “Are we equipped to practice it with equity?” 

What ‘AI equity’ really means (and why it’s hard to live it) 

AI equity refers to the ethical development, deployment, and use of artificial intelligence—ensuring AI systems prioritize fairness, inclusivity, and justice for historically marginalized communities. It’s rooted in data equity, which focuses on how data is collected, interpreted, and governed; whose stories are amplified and whose are erased. In other words, data equity is the foundation; AI equity is the future structure built upon it. A structure built upon an unjust foundation will only scale that harm. This means AI equity is less about tools and more about governance, participation, and care. 

AI awareness is growing, but readiness is falling behind 

The 2025 findings show nonprofit staff learning about AI and data equity faster than ever, but implementation of equity practices dropped significantly. Several forces seem to be converging at once: 

• Nonprofits are experimenting faster than they’re building guardrails. 
• Only about 15% currently have an organizational policy for responsible AI use. 
• Staff interest is high, but time, training, and budgets have not kept pace. 

 The sector’s increasingly familiarity with the right vocabulary reflects an aspirational stage of growth—attempting to name what they value while waiting for the infrastructure needed to make it real. 

As one respondent reflected in the report, “The sector has momentum, but it needs scaffolding.” In other words, the learning is there; the structures are not. 

The structural gaps behind the decline in AI equity practices 

The drop in the percentage of organizations implementing equity practices is not rooted in unwillingness but in limited capacity. Qualitative responses from the research point to three major structural constraints: 

• For nonprofits: Especially at small and midsize organizations, staff lack the time and internal process support to translate learning into governance. 
• For funders: Many are still funding experimentation, innovation, or tools but not governance, policy development, or staff time to learn. 
• Across the sector: The speed of AI development is outpacing the frameworks and protections required to adopt it responsibly. 

In short, nonprofits are not ignoring equity. They are being outpaced by the velocity of change around them. 

Moving from awareness to practiced AI equity 

So, what would it look like to close the gap between knowing and doing? Here are some ways organizations can put the idea of AI equity into practice. 

• For nonprofits: Embed equity questions into every AI decision—who benefits, who could be burdened, whose voice is missing, and what risk is possible? 
• For funders and philanthropic intermediaries: Fund governance and learning capacity, not just integrating AI tools into workflows or pilot adoption. Policy and people need support before platforms. 
• For sector ecosystem builders: Prioritize shared learning models—such as AI Readiness Studios—so nonprofits aren’t learning in isolation or competing for knowledge. 

Practiced equity is not a template or checklist locked in our systems. It is a discipline—built slowly, relationally, and iteratively. 

The AI Equity Project’s 2025 findings reveal a sector in the moment of tension: one that’s increasingly aware of the importance of AI equity but not yet fully equipped to implement it. Readiness to integrate AI tools while ensuring equity requires more than interest; it requires the conditions to act with care: the time, resources, and training to develop policies and practices. 

AI equity, then, isn’t a milestone to reach but a practice to sustain. It’s not proven by using the right vocabulary but by building the structure needed to ensure AI systems prioritize fairness, inclusivity, and justice for marginalized communities. And as technology evolves, the question will not simply be whether nonprofits can keep up but whether the systems around them allow equity to be more than an aspiration. 

Photo credit: MTStock Studio/Getty Images