30 minutes • Establish frameworks for responsible AI use and funder confidence
Trust isn't automatic. Trust is earned through consistent demonstration of competence, honesty, and reliability. For AI use in grants, trust comes from transparency about how you use AI, quality controls you maintain, and results you achieve. Transparent organizations build trust faster than secretive ones.
Complete trust framework includes: clear policies on when and how AI is used, quality assurance processes ensuring accuracy and appropriateness, staff training ensuring responsible use, governance ensuring oversight, reporting enabling visibility, and results demonstrating value. Comprehensive framework builds trust with funders and staff alike.
Work with staff, leadership, and ideally funder representatives to develop clear policies. What is AI used for? What are explicit limitations? How is quality assured? How is use disclosed? When is AI not used? Inclusive policy development creates buy-in and comprehensive guidelines. Policies shouldn't be constraints—they should enable responsible use.
Policies might cover: approved tools (which AI systems can staff use), use cases (what tasks AI can assist with), limitations (what AI shouldn't do), verification requirements (what must be checked before using), disclosure requirements (when to tell funders), staff training requirements, quality standards (must meet same standards as non-AI work), and override procedures (when humans override AI).
Policies are only valuable if people know and follow them. Publish policies. Train staff on them. Make them easily accessible. Include in onboarding. Communicate them to funders if relevant. Clear, published policies demonstrate responsibility.
When AI generates content (draft section, research findings, analysis), it must be verified before use. Verification might include: accuracy checks (are facts correct?), appropriateness checks (does it fit organizational voice and values?), completeness checks (does it address all requirements?), consistency checks (does it align with other content?). Document verification processes.
Establish non-negotiable human checkpoints. Strategy decisions? Always human. Accuracy of facts? Always verified. Final approval before submission? Always human. Humans remain in critical roles; AI assists but doesn't replace judgment.
Track quality: Do AI-assisted proposals have same funding rate as human-generated ones? Are quality scores comparable? Do funders report concerns? Monitor metrics. If quality drops after AI adoption, adjust approach. Quality metrics prove to funders that AI isn't compromising work.
Create a simple checklist for verifying AI output: Facts verified against authoritative sources? Voice consistent with organizational norms? Tone appropriate? Content addresses all requirements? Proposal instructions followed? Checklists ensure consistency and provide evidence of verification to funders if questions arise.
Share evidence: "We've used AI in 50 proposals this year. Acceptance rate: 32% (compared to 28% previously). Funder feedback on proposal quality: positive." Numbers convince skeptics. Data shows AI is improving or maintaining outcomes, not degrading them.
Document specific examples: A proposal that benefited from AI research. A workflow that was streamlined. Time saved without quality loss. Case studies make abstract benefits concrete. Share (with appropriate confidentiality) specific examples of value created.
Keep records showing: when AI was used, what for, how output was verified, who approved. These records aren't just for internal use—they document responsible use. If a funder audits or questions, documentation shows you were thorough.
Establish clear governance: Who makes decisions about AI use? Who approves new tools? Who handles concerns? Who reports to leadership? Clear governance demonstrates that AI use isn't ad-hoc but systematized. Shared responsibility (not one person deciding) strengthens governance.
Monthly review of AI use: Are guidelines being followed? Are quality metrics showing positive trends? Are staff concerns emerging? Oversight catches problems early. Transparent oversight also demonstrates responsibility to stakeholders.
What happens if AI use causes problems? If a funder complains? If quality issues emerge? Clear escalation procedures show you can handle issues responsibly. Problems happen; how you handle them reveals character. Transparent, rapid response to problems builds trust.
When you invest in training staff on responsible AI use, that investment demonstrates seriousness. "Our team completed AI competency training" shows commitment to responsible use. Training documentation is evidence of responsibility.
Consider and address ethical questions: Does AI use cause unintended consequences? Are there fairness issues? Data privacy concerns? Addressing these proactively shows sophisticated thinking. "We've considered these ethical questions and here's how we address them" is powerful message.
If a funder values equity, show how AI helps (research on equitable approaches) and doesn't harm (quality and accuracy protected). If funder values innovation, show AI as innovation. Demonstrate that your AI use aligns with funder values and furthers their goals.
Do what you say. Say you verify AI output? Verify it. Say you track metrics? Track them. Consistency builds credibility. Inconsistency destroys it. Long-term credibility comes from years of consistent, reliable behavior.
When funders raise concerns, respond seriously and quickly. "We've heard your concern about AI accuracy. Here's what we're doing." Responsiveness shows accountability. Defensive or dismissive responses damage trust.
AI is one tool among many. Build trust broadly in your competence, reliability, results, values alignment. Strong relationships survive questions about individual tools. Trustworthy organizations gain benefit of the doubt when tools are questioned.
Some funders might prohibit or restrict AI. Respect their restrictions. Communicate understanding: "We respect your preference. We won't use AI on this grant." Respectfulness to funder requirements builds trust even when you disagree.
If a funder questions your AI use, respond with transparency, evidence, and openness. Share your governance. Show your quality assurance. Present data. Welcome conversation. Defensive responses raise suspicion; open dialogue builds understanding.
If AI use causes a problem (inaccuracy discovered, funder expresses concern), address immediately. Acknowledge. Explain what happened. Share what you're doing to prevent recurrence. Responsible handling of problems actually strengthens trust.
A nonprofit proactively created and shared their AI use policy with all funders. "Here's how we use AI responsibly." Most funders responded positively. A few had questions, which the nonprofit answered. One funder restricted AI use on their grant; nonprofit respected it. Proactive transparency created trust across the board.
Trust building is continuous. Maintain transparency. Update policies as tools and understanding evolve. Share results and learnings. Stay engaged with funders. Over time, as AI becomes normal and benefits are demonstrated, trust deepens. Organizations leading thoughtfully on this issue will have advantage.
Next, we'll explore creating documentation and reports that demonstrate AI use and impact to funders.
Continue to Next Lesson