Responsible AI
You trust us with information about your organization, your funders, and your programs. Here is exactly how we use AI, what we don't do with your data, and how we hold ourselves accountable.
Our Commitment
The Almoner, LLC — the company behind AI Grant Helper — is a signatory of the Fundraising.AI Framework for Responsible and Beneficial AI. This framework governs ethical AI use in nonprofit fundraising, a sector representing nearly $500 billion annually in the United States.
We commit to 10 principles: Privacy & Security, Data Ethics, Inclusiveness, Accountability, Transparency & Explainability, Continuous Learning, Collaboration, Legal Compliance, Social Impact, and Sustainability. This is not a marketing badge. It shapes every product decision we make.
How AI Grant Helper Uses AI
"AI" appears in our name, so you deserve a clear accounting of what that means.
Foundation & Grant Search
Not AI. Search uses standard database queries and filters. When you search for foundations by state, assets, or focus area, you are querying structured IRS data and Grants.gov records — not a language model. Results are deterministic: the same search returns the same results every time.
Foundation Enrichments
AI-generated, human-reviewed. The program focus, geographic reach, application guidelines, and contact information shown on foundation detail pages are generated by AI models analyzing IRS filings and foundation websites. This data is generated offline, reviewed for quality, and stored as static records — no AI runs when you view a foundation page.
Funder Matching
Rule-based scoring, not machine learning. When we recommend funders for your organization, the scoring uses weighted rules built from 14 years of fundraising experience: geographic overlap, grant size fit, sector alignment, and giving activity.
Planned: Writing Support Tools
Coming later. Future features — like budget structuring, proposal feedback, and funder fit analysis — will use language models (Claude, Gemini) to assist your writing process. These tools will help you draft and refine, not write for you. Every planned AI feature will include a clear explanation of what the model does and what you should verify.
What This Means for You
Your data stays yours
Your organization profile, saved funders, and application history are yours. We do not sell your data, share it with other platforms, or use it to train AI models. When we use AI services to power features, your data is sent only for that specific request and is not retained by the AI provider for training.
AI does not write your grants
We believe funders deserve to hear your voice, not a machine's. AI Grant Helper is a research and management tool. When we add writing features, they will structure, suggest, and check your work — not replace it. Your proposals should sound like your organization because they come from your organization.
We tell you when AI is involved
Every feature that uses AI is labeled. If a recommendation, summary, or suggestion comes from a language model, you will know. If it comes from IRS data or your own records, you will know that too. No black boxes.
We warn you about AI limitations
AI can produce convincing text that contains fabricated statistics, incorrect citations, and wrong information about your funder. Every AI-powered feature includes clear guidance: treat AI output as a starting point, not a finished product. Verify every fact before submitting anything to a funder.
We respect dignity
The Fundraising.AI Framework prohibits exploitation, manipulation, deception, and coercion in AI-driven fundraising. Our tools will never help generate manipulative appeals, exploit beneficiary stories, or pressure donors. This aligns with Catholic Social Teaching on human dignity — it is mission, not merely compliance.
We build for small budgets
Our partners operate on minimal budgets. We choose the right-sized technology for each task, design for provider portability, and keep AI costs proportional to value delivered. A tool that becomes too expensive for the smallest community organization has failed.
Report a Concern
If you believe any of our tools or practices fall short of these commitments, I want to hear about it directly. Contact Nathan Krupa at nathan@thealmoner.com.