Advice From 119 AI-Powered Nonprofits: Inside Fast Forward's 2026 AI for Humanity Report
Fast Forward's survey of 119 AI-powered nonprofits finds AI is making their services more efficient but most can't fund the jump past the pilot, and its first advice to every nonprofit is to start from a real community need before choosing AI.
Best for: Executive directors, program leads, development and grants teams, nonprofits planning to use AI in their programs or services
Don't adopt AI because it's available. Don't adopt it because a funder is offering money for it. Start with what the people closest to the problem need. Then ask whether AI is the right tool.
That's the first recommendation for nonprofits in Fast Forward's 2026 AI for Humanity Report. It draws on a survey of 119 AI-powered nonprofits across 20 countries. These are organizations that use AI to deliver their programs and services, not only to run the back office.
The report calls them a growing subset of nonprofits, and argues they can serve as models for how other organizations think about AI's risks and rewards. Here's what they reported, and the advice worth borrowing.
Details
Who Fast Forward surveyed
Fast Forward supports nonprofits building tech and AI, and publishes research on the wider sector. For this study it partnered with Third Plateau, an impact consultancy that designed and ran the research. Google.org supported the report. Fast Forward shared the highlights on September 15, 2026.
- Who: 119 AI-powered nonprofits (APNs), plus 20 interviews with nonprofit leaders and funders
- When: surveyed in spring 2026
- Worth knowing: the findings are self-reported and don't represent the whole nonprofit sector. 46% of respondents were part of Fast Forward's own portfolio, and nonprofits using AI only for internal operations weren't included.
What they reported
- 92% say AI made their service delivery more efficient, and 55% say it made personalized services at scale possible
- 90% of those piloting or scaling a solution have a plan to scale it, but only 24% have the resources to carry it out
- 45% name rising costs of AI tools and infrastructure as the top external threat to their AI work
- 88% are concerned about AI delivering inaccurate information to the people they serve
Fast Forward is careful here. These are participants' own perceptions, and the report says they shouldn't be read as proof that AI caused the changes.
Why this matters for your nonprofit
APNs are still a subset of the sector, but some of what they learned travels well. Asked what enabled AI adoption, 64% named a champion who drives the vision, 45% a culture of experimentation and willingness to fail, and 42% curious staff. All three are about people, not technology.
Their advice is most useful when your team considers putting AI into a program, a grant proposal, or anything that touches the people you serve. Not every recommendation fits every team. The one about contributing to a shared commons of open-source tools and data matters most if you build your own AI.
How to get started
Steps
- Start with the need, not the tool. Before your next AI idea goes further, write down the problem the people closest to it have named. 39% of surveyed APNs called a clear problem to solve a top-three factor in adopting AI. Adalat AI is one example: it built a stenography tool for a real pain point in India's courtrooms, and its tools now run in 5,500 courtrooms across 11 states.
- Decide what stays human, and say so. The report's advice is to use AI to build better processes but keep decision-making human, and to keep disclosing your AI use to users. Two-thirds (66%) of surveyed APNs worry AI could erode trust with the people they serve.
- Price the hidden labor. If AI is part of a grant proposal, name the line items for maintenance, evaluation, and iteration so funders see what the work costs. 28% of surveyed APNs say AI cost more than they expected. Your grants lead is the natural owner.
- Build in community feedback. 76% of surveyed APNs collect feedback on their AI tools after launch, but only 28% seat community members on an advisory or governance board. The report urges involving the people your tools serve in shaping them.
Pro tip: Borrow the report's language with funders. Interviewees asked them to pay attention to the "messy middle" between pilot and proven solution, where costs rise with use.
The bottom line
Start with the need. The tool comes second. Reading the report is the easy part. The bigger cost is the one it keeps coming back to: once AI is in use, "monitoring, testing, and iteration are where the money actually goes."
The nonprofits in this report didn't start with the technology. They started with a need their community named. For them, AI changed how they deliver, not what they're working toward.
References
- The AI Playbook’s summary of the report to quickly review key highlights
- AI for Nonprofits: Findings from the 2026 AI for Humanity Report: Source article, Fast Forward, September 15, 2026
- Two-page summary of Fast Forward's 2026 AI for Humanity Report
- 2026 AI for Humanity Report: Full report, with methodology and recommendations for nonprofits and funders
- Fast Forward: Supports nonprofits building tech and AI, and publishes research on the sector
- Third Plateau: Impact consultancy that designed and conducted the study
- Google.org: Supported the report
- Adalat AI: AI-powered nonprofit example from Fast Forward's highlights
Source: Fast Forward (2026 AI for Humanity Report, research by Third Plateau), September 15, 2026. Read on AIPlayBook.md, a Fall Communications resource. Every fact above comes from the source linked here.