What is an AI governance policy for a nonprofit?

An AI governance policy sets the standards a nonprofit follows when staff and partners use AI. It covers privacy and security, data ethics, accountability, transparency, and legal compliance, and it names who decides on new uses of AI. Empowered Fundraiser treats governance as a practice, so the policy gets reviewed as tools and uses change.

What does the Empowered Fundraiser AI policy cover?

The policy, A Framework for Responsible AI in Client Partnerships, sets ten principles. They are continuous learning, privacy and security, data ethics, inclusiveness, accountability, transparency and explainability, collaboration, legal compliance, social impact, and sustainability and environmental responsibility. Empowered Fundraiser holds itself and its clients to these as minimum standards.

How do I get a copy of the AI policy?

Enter your first name and email in the Get Our AI Policy form on this page, and the policy arrives at no cost. You also receive Fundraising NOW, the Empowered Fundraiser newsletter, and you can unsubscribe anytime.

What does responsible AI consulting for fundraising include?

Responsible AI consulting covers four parts. Anne assesses where AI fits, teaches staff to use it well, writes the policies that keep it accountable, and implements it. The scope reflects the organization’s responsibilities and the uses of AI it is considering.

Can Anne help us evaluate AI tools and vendors?

Yes. Anne evaluates where AI fits in your fundraising operation and coordinates implementation with your existing vendors. When no existing tool fits a defined need, the work can include building a bespoke tool.

How does staff education support responsible AI adoption?

Staff education gives your team the judgment to make sound decisions about AI in daily work. It works alongside policy and oversight, so the organization keeps both the capability and the accountability in-house.

Anne Murphy is CEO of Empowered Fundraiser Consulting and an AI operations consultant specializing in AI governance, education, and responsible AI adoption. Anne co-authored the Fundraising.AI framework for responsible AI in fundraising and has educated more than 4,000 professionals in responsible AI.

Nonprofit AI Policy and Responsible AI in Fundraising

A Framework for Responsible AI in Client Partnerships

As a global leader in responsible AI for nonprofit fundraising, Empowered Fundraiser Consulting takes the use of generative AI in any and all knowledge work seriously. We advocate in international, national, and regional settings for AI to be rolled out with care.

In our collaboration, our policies apply to clients as the minimum standards for responsible AI.

Misuse of AI - including trying to avoid it - puts us all at risk. AI is in every consumer and enterprise platform most people have access to. Learning to navigate the risks and rewards is a responsibility we share collectively.

1. Continuous Learning

EFC and our clients commit to staying informed about responsible AI and applying what we learn in our work together.

At our launch meeting, we will all share the technology systems and datasets we use in our work (client and EFC), discuss the policies and systems that govern their use, and settle on an agreed technology stack and data environment.

2. Privacy and Security

EFC and the client protect personal and sensitive information and comply with applicable data-protection laws. Both parties respect the privacy of donors, clients and colleagues.

Consent. Address consent as appropriate and explain the intended use.
Control over the use of data. Respect the permissions and limits attached to the information.
Ability to restrict data processing. Support the mutually agreed-upon route for requests to limit how information is processed.
Right to rectification. Support one another’s route to correct inaccurate personal information and address affected records or decisions.
Right to erasure. Support the client in handling deletion requests and explain applicable retention requirements or other limits.
Right of accessibility. Make applicable access routes and information about data understandable and usable.
Adherence to privacy laws. Follow the highest standard of relevant privacy requirements, alongside safeguards appropriate to our role and the information involved.

Passwords, government identification numbers, Personally Identifying Information, financial account details and health records belong only in tools specifically approved to handle that information. These are the most restricted safeguards.

EFC and our clients use agreed storage and sharing arrangements. We keep client information separate and limit access to the people who need it. Both parties will use only the tools and sharing arrangements agreed for our work. We ask clients to tell us about restrictions on their data and to obtain agreement before sharing EFC materials or project information elsewhere.

3. Data Ethics

EFC and our clients commit to ethical data practices, ensuring the data we use is accurate, relevant and collected with proper consent. Both parties respect the purposes for which information was shared and the limits on its use. Our work must preserve people’s dignity and their ability to make informed choices about their information.

4. Inclusiveness

EFC applies its AI and fundraising expertise to identify and address bias in the tools, data and outputs we use. We ask clients to contribute their knowledge of their organizations and communities as we examine who is represented and how AI influences decisions.

5. Accountability

We share accountability with our stakeholders for the AI applications we develop and utilize.

EFC takes responsibility for the AI-assisted work we deliver, including its accuracy and its effects on clients and donors. We verify substantive findings, exercise professional judgment, and correct errors in the work and any recommendations based on it. We expect clients to apply the same care to AI-assisted work they contribute to the engagement and to take responsibility for its accuracy and effects.

6. Transparency and Explainability

We explain how AI contributes to our work and how we review its results. We keep clients informed when our use of AI changes in ways that affect them. We ask clients to be equally open about how AI contributes to work they share with EFC and to tell us when that use changes in ways that affect the engagement.

7. Collaboration

We engage with peers to share experiences, challenges and successes with responsible AI, and encourage our clients to do so as well.

EFC is a global leader in responsible AI for nonprofit fundraising. We’ve educated over 4,000 colleagues and technology providers - including frontier labs like OpenAI.

We encourage clients to share what they are learning during their work with us with their colleagues.

8. Legal Compliance

EFC and the client each comply with the laws, regulations and contractual obligations that apply to their work, including AI, data protection and fundraising.

9. Social Impact

We aim to maximize the positive social impact of AI while minimizing harm.

We consider how AI affects the people and communities our clients serve. Respect for dignity and individual choice guides our decisions.

EFC and the client will consider how AI changes staff responsibilities and the skills teams need. We encourage clients to discuss these changes with their teams and make decisions that reflect their organization’s values and the people affected.

10. Sustainability and Environmental Responsibility

We consider the long-term sustainability and environmental impact of the AI we use, and advocate for sustainable AI practices within EFC and the broader fundraising community.

EFC and the client will consider energy and water use, equipment waste and effects on communities when choosing AI tools for the engagement. We encourage clients to include these considerations in their wider technology decisions.