Harnessing AI for Good: Streamlining Nonprofit Operations

Par Dave Norris , cofondateur de Bold Crow AI et PDG de Proofpact.
04 juillet 2023

Alongside their passionate goals and tireless pursuit of their missions, nonprofits often grapple with the daunting task of managing complex administrative processes. We have all heard and used the phrase « work smarter, not harder » at one point or another. Enter the world of artificial intelligence (AI), a technology revolutionizing workflows and alleviating many of the burdens of administrative work.

In most cases, I would be a proponent of going against the grain and saying that nonprofits should not attempt to do more with less. The overhead myth is just that, a myth. In truth, nonprofits should strive to place emphasis and focus on what makes them operationally successful, just like any other organization. What I will say though is that AI can play a critical role in this just as it is in the for-profit world.

My next and last disclaimer is that I want to stress that AI should not replace humans. As humans, we will evolve with it just like we did Web 2.0 and like we are doing with Web 3.01. AI should be used to augment human workflows, improve automations, and enhance decision-making processes. AI should be seen as a valuable tool to augment human workflows and empower nonprofit professionals.


AI Tools and When to Use Them


There are a few popular AI tools out there. You have likely heard of some of the generative AI tools such as OpenAI’s ChatGPT or Google Bard for example. There are plenty of good individual use cases for tools like ChatGPT. However, there are many instances in which ChatGPT should not be used too. Below are a few examples for each.

When to Use ChatGPT:

  1. Generating personalized communications: ChatGPT can be used to draft personalized thank-you emails or donor acknowledgment letters.
  2. Assisting with grant application research: ChatGPT can assist nonprofit professionals in the initial stages of grant applications by gathering information on eligibility criteria, funding priorities, or past recipients.
  3. General summarization: ChatGPT can be used by nonprofit professionals to summarize large amounts of copy to more quickly extract the details they need.

When Not to Use ChatGPT:

  1. Handling sensitive donor information: ChatGPT should not be used to process sensitive donor or personal data directly. As a publicly available tool, it is not designed to meet stringent data privacy and security standards.
  2. Making high-stakes decisions: While ChatGPT can offer insights, it should not be solely relied upon for critical decision-making. Complex decisions that involve legal, financial, or ethical implications should involve human expertise and judgment.
  3. Providing health or legal advice: ChatGPT is not a substitute for professional advice in highly regulated fields like health or law. It lacks the specialized expertise required to provide accurate and reliable guidance.

Remember, ChatGPT and similar generative AI tools should be seen as valuable assistants rather than authoritative sources. They excel at automating routine tasks and generating initial drafts, but human oversight and expertise are essential to ensure accuracy, empathy, and compliance with industry standards and regulations.


AI Agents and Large Language Models

Large language models (LLMs)2, have made significant strides in natural language processing, enabling applications to more precisely and quickly iterate on tasks that involve generative text-based iterations. This has opened the door to something called AI agents. AI agents are the interface between LLMs and external resources.

Agents work on your behalf and complete tasks using tools (functionality) with the objectives that you give them and can do so autonomously or through interactive steps with a user. Tools can be things like access to web searches, connections to a database, or custom functions for AI agents. One example might be that an agent has access to Google search so that it can search the web for weekly news. Another example might be that the agent has access to a tool that allows it to more accurately run mathematic equations.

These agents make calls to the LLM when needed and can use complex reasoning to choose between an array of tools when they deem necessary. The instructional prompt can be complex, contain numerous objectives, involve multiple tools, and encompass side chained prompts that initiate the agent’s actions.


AI Agents for Nonprofit Admin Operations

Now, let’s dive deeper into the transformative power of AI agents and how they can revolutionize nonprofit administrative tasks. By leveraging AI agents, nonprofits can automate and optimize various aspects of their operations, freeing up valuable time and resources to focus on their core mission.

1. Enhancing Data Management and Analysis

AI agents can:

    • Make data more accessible by providing a way for entire teams to query the database with natural language.
    • Help standardize and cleanse data that may be relatively “dirty” and help make it more reliable.
    • Extract relevant information from documents, forms, and emails, reducing manual data entry efforts and ensuring data accuracy.
    • Analyze data patterns, donor trends, and program outcomes, offering valuable insights for strategic decision-making and resource allocation.

2. Streamlining Volunteer Coordination

AI Agents can

    • analyze volunteer skills and availability, matching them with appropriate projects and reducing the administrative burden of manual coordination.
    • Guide volunteers through the onboarding process, providing them with essential information and resources, while also collecting necessary documentation.
    • Provide a centralized platform for volunteers to communicate, ask questions, and receive updates, creating a seamless flow of information.

3. Enhancing Donor Management and Engagement

AI agents can:

    • Analyze donor data and generate personalized emails, newsletters, and updates to foster stronger connections with supporters.
    • Track donor interactions and automate timely follow-up messages or conversations, ensuring donors feel valued and appreciated.
    • Analyze donor behavior and preferences to provide insights for targeted fundraising campaigns, leading to improved donation conversion rates.



The integration of AI into nonprofit operations holds immense potential for driving efficiency, effectiveness, and mission success. By harnessing the power of AI responsibly, nonprofits can streamline administrative processes, enhance admin productivity, and ultimately enable nonprofits to focus more on their core mission. The benefits of utilizing AI agents in areas such as donor management, volunteer coordination, data analysis, and others are clear.

While the integration of AI into nonprofit operations brings numerous benefits, it is essential to emphasize the importance of maintaining a human component in donor relations. AI can undoubtedly support donor relation strategies by automating certain tasks, such as identifying patterns in donor behavior, writing first draft appeals, and providing reminders for outreach. However, it should not replace the personal touch and genuine human connection that is crucial for fostering strong relationships with donors. AI is not meant to diminish human interaction. When used correctly AI augments and enhances authentic human connection.

As we look to the future, nonprofits should actively seek opportunities to ethically integrate AI into their operations, empower their teams with AI agents, and encourage their staff to responsibly leverage tools like ChatGPT to achieve greater efficiency and effectiveness. By embracing AI, nonprofits can embark on a transformative journey that revolutionizes their processes and helps them focus on what truly matters – creating positive change in the world.



Dave Norris is a former digital agency owner, now co-founder of Bold Crow AI, and CEO of Proofpact.
He helps businesses and nonprofits implement customized AI solutions in meaningful, responsible, and ethical ways.

Cet article fait partie de l’édition spéciale de juillet 2023: Philanthropie et intelligence artificielle. Vous pouvez trouver plus d’informations ici.

Philanthropie et Intelligence artificielle


[1] The term « Web 2.0 » emerged in the early 2000s to describe a shift in internet technologies and user interactions. It represented a transition from static web pages to dynamic platforms that enabled user-generated content, social networking, and collaboration. The term « Web 3.0 » or the « Semantic Web » refers to the envisioned next phase of the internet, characterized by decentralization and intelligent data-driven services and enhanced machine-to-machine communication.

[2] A large language model (LLM) is a computerized language model consisting of an artificial neural network with many parameters (tens of millions to billions), trained on large quantities of unlabeled text using self-supervised learning or semi-supervised learning.