By Rachel Qian
AI is quickly becoming part of everyday work in nonprofit organizations. The most important goal now is to help everyone in the organization use AI confidently and responsibly.
The Mistakes Nonprofits Don’t Realize They’re Making With AI
When people hear “AI safety,” they often imagine dramatic scenarios: autonomous systems making harmful decisions or sophisticated cyberattacks.
While those can be possible, most nonprofits face less aggressive yet more common challenges.
1. Treating AI as an Individual Tool Instead of an Organizational Practice
Without guidance, every employee develops their own way of using AI.
Some use it to brainstorm ideas. Others draft communications or summarize research. Each usage may be reasonable on its own, but without shared expectations, it’s difficult for teams to establish consistent safety practices.
2. Lack of Guidance on What Information Goes Into AI Tools
AI often produces better results when it has more context, which can make it tempting to paste in entire documents, spreadsheets, or conversations. However, not all information should be shared with every AI tool.
Organizations should take a little time to identify what types of information are appropriate to use with AI and what should remain internal. For example, public-facing content or general brainstorming may be perfectly appropriate, while donor records, client information, financial data, or confidential board discussions may require additional safeguards or should only be used with approved tools.
3. Waiting Until There’s a Problem
Many organizations assume they’ll create AI guidelines once adoption becomes more widespread.
Organizations that benefit the most from AI aren’t necessarily the ones using the newest tools. They’re the ones that establish simple expectations early, making it easier for staff to adopt AI with confidence as new opportunities emerge.
The Thoughtful, Non-Overwhelming Ways to Adopt AI
The good news is that responsible AI adoption doesn’t require a committee, a consultant, or a hundred-page policy. For most nonprofits, a few practical habits make an enormous difference.
1. Keep Humans Responsible
Keeping humans in the loop is the first and most important approach. AI is a powerful collaborator, but people should remain responsible for the organization’s decisions and communications.
For example, if the organization is using AI to draft grant proposals, a team member should always review the final work before it’s shared for review or submitted.
Keeping people in the loop helps organizations benefit from AI’s speed while ensuring that important decisions continue to reflect human judgment.
2. Choose Trusted AI Tools
Not all AI tools handle information in the same way. Before adopting a new AI tool, spend a little time understanding how it stores and uses data. You don’t need to perform an extensive technical review, but understanding a tool’s data policies can help your organization make informed decisions about when and how to use it.
Understanding how every AI tool handles data can feel overwhelming, especially for nonprofits with limited time and technical resources. Fortunately, you don’t have to read every privacy policy line by line. AI can help with that, too.
Before adopting a new tool, ask your preferred AI assistant to summarize the tool’s privacy policy, terms of service, or trust center. Questions like “How does this tool use my data?” or “What types of data should I avoid uploading?” can quickly surface the information that matters most. While it’s still worth verifying important details, AI can significantly reduce the time needed to evaluate whether a tool is appropriate for your organization.
3. Establish Boundaries
One of the most effective things a nonprofit can do is create a simple and concise AI usage guide. It just needs to provide enough direction for people to make good decisions.
For example, your guidance might answer questions such as:
· What types of tasks is AI encouraged to help with?
· What information should never be entered into public AI tools?
· When should AI-generated work be reviewed by another person?
· Which AI tools has the organization approved for use?
Having a few shared expectations helps everyone approach AI with greater confidence and consistency, without creating unnecessary bureaucracy. Try this free template for your organization’s AI Policy.
AI Safety Starts with Small Steps
Responsible AI adoption doesn’t require perfection. It starts with a few thoughtful decisions that help people use these tools with confidence.
By keeping humans involved, choosing trusted AI platforms, and establishing a handful of shared guidelines, nonprofits can create an environment where AI becomes a reliable partner rather than an uncertain experiment.
Rachel Qian is Director of Product & Design at PlanPerfect, where she leads product strategy and design. She holds an MBA from Kellogg School of Management and brings a diverse background in product design, having previously worked at Slickdeals and run her own design consultancy. Since 2024, Rachel has focused on building AI-driven products and experiences, bringing a human-centered perspective to the challenges of creating safe and intuitive AI systems. Outside of work, she has volunteered with several nonprofits in Los Angeles and Chicago.
