How to use nonprofit AI tools effectively and ethically in 2024

The use of AI has become a hotly contested issue across industries—including the philanthropy sector. In fact, according the recent webinar, “Using A.I. to Streamline Nonprofit Operations” hosted by The Chronicle of Philanthropy, research suggests more than half of nonprofits now use generative AI for at least some daily tasks, such as content creation.

Given this growing usage, nonprofit leaders are struggling with some pretty huge questions: Do AI and machine learning belong in the nonprofit world? How is AI changing nonprofits? And what are the ethical considersations of using it?

The Chronicle of Philanthropy’s webinar, which featured several nonprofit experts as panelists, discussed some of these exact questions about AI for nonprofits, including how organisations might use machine learning effectively and ethically for years to come.

The challenges surrounding AI for nonprofits

The adoption of any new tech into an organisation can be overwhelming. When it comes to the introduction of AI, the same is true. According to research, some of the top challenges nonprofits face when implementing AI are:

Familiarity: More than three-fifths of nonprofits, according to COP, said they lack familiarity with AI, which makes it difficult to find uses for it.

Resources: Another half surveyed said that finding enough funding and training for staff to understand these tools were significant challenges, too.

Guiding principles, rules and considerations: Nearly 80% of nonprofits still lack an organisational-wide policy for AI that outlines how to use the tools ethically and safely, leaving many feeling uncertain about their role.

How nonprofits can use AI

Given these challenges, panelists discussed a few potential solutions and thoughts to consider on how to implement AI effectively and ethically so that everyone in an organisation feels confident and comfortable.

Encourage curiosity

Given the unfamiliarity that so many nonprofits note, leaders should take the time to encourage and invest in becoming comfortable with AI tools in a few different ways:

Experiment responsibly: Select a few AI tools and start experimenting with low-risk tasks. Use AI to consolidate data, produce content drafts and the like, just to get a feel for how they work and assess their strengths and weaknesses.

Encourage a culture of curiosity: Nonprofit leaders should take the initiative to create a culture within their organisation to supports experimentation. One of the best ways to do this is by setting an example and experimenting themselves.

Offer tangible tools and resources: Don’t just leave your team up to their own devices (no pun intended). Pre-select a handful of tools and carve out time for everyone to experiment. Or, carve out an innovation budget that allows staff members to purchase or download possible tools.

Understand AI’s strengths and weaknesses

Understanding the abilities and limitations of AI and machine learning for nonprofits is one of the most important parts of integrating it ethically and responsibly into your organisation. When introducing it to your operations, acknowledge that AI tools should:

• Be used for menial tasks: AI can be thought of as a co-pilot or an assistant. These tools can help take care of mundane, repetitive tasks so nonprofit staff members can spend more of their time and energy on driving impact and and moving the mission forward. These tasks might include using AI for grant applications, content synthesisation and the like.
• Not be used for original thought: AI is not equipped for creativity or original thinking, meaning the tools shouldn’t be used for mission-critical tasks like narrative change and storytelling.

Create and abide by organisational guiding principles

One of the most essential components of using AI in your nonprofit is using it responsibly and ethically. While experimenting and understanding its strengths and weaknesses are important, developing guiding principles for implementation is crucial, too. To do so, follow these steps:

Step 1: Determine the role that AI will play in your operations

As Nick Cain of the Patrick J. McGovern Foundation noted, there’s a correlated relationship between the level of scrutiny that an organisation should place on an output created by generative AI, and the proximity of that output to an actual beneficiary.

This means that nonprofits that intend to use AI in various mission-critical tasks or tasks that will directly impact beneficiaries should be much more stringent about their guiding principles and expectations of using these tools. Nonprofits that use AI for content synthesisation, for example, might engage in less scrutiny.

Step 2: Get your organisation on board

Consider the areas that your staff is telling you they’d like to use these tools, and then create a high-level list based on reported needs.

Then, start to think about which tools everyone feels comfortable using or not using. Cain  recommends taking a “stoplight” approach to categorising specific AI tools:

• Green: Tools labeled as green indicate staff’s higher level of comfort using the application and a willingness to integrate it into operations.
• Yellow: Yellow indicates a middle range of comfort, which might include staff’s openness to experimentation, or need for expert guidance on implementation.
• Red: Tools labeled red indicate the least amount of comfort, flagging the tool as one that the organization does not want to use.

Step 3: Define parameters for confidentiality, security, and transparency

Create a policy that underlines how crucial it is never to put private or confidential information about the program or your participants/beneficiaries into the tools.

Plus, consider how you want to handle transparency. Consider whether you’ll screen for AI content, or if you’ll inform funders, donors or other stakeholders about the use of machine-learning tools.

Using AI for good

A lot of nonprofit leaders understand that machine learning is here to stay, so getting comfortable with AI tools—and figuring out how they do or do not fit in with your mission and operation—is essential to long-term success.

While there’s not one right way to integrate AI, be sure to involve your team, seek guidance from experts and take the time to learn.