Beyond the Buzz: A Q&A on How Nonprofits Can Best Leverage AI

Artificial Intelligence (AI) has been around for over a decade, but in 2024 it had its breakthrough moment, with discussions about its impact touching every industry, including the nonprofit sector. I first learned about AI at DevLearn in 2019, which inspired our 2020 trends blog about technology.

Since then, I have been trying out AI for the past few years for various tasks. I no longer turn out the lights to go to bed?—?I just tell Alexa to do it. I ask ChatGPT to add some extra pizzazz to speech titles. Google AI helps me kickstart tailored birthday greetings to friends and family members.

But there is a darker side?—?I have had to battle students on the use of AI on quizzes and have gotten those weird reply emails from colleagues that were obviously written by AI. I even tried Insta HeadShots once for fun, but the results were terrible. Let’s just say I’ll be hiring a photographer to take a new one when the time comes.

To explore how nonprofits can navigate AI, I joined forces with one of my favorite colleagues, Nicola Paugh, Ph.D., Executive Director of The Nonprofit Alliance of Bermuda, to break down what AI means, how Generative AI differs from Traditional AI and how nonprofits can use it without losing authenticity.

What is AI? How is Generative AI different? What are the practical implications? How are real people using it to be more productive?

Traditional AI analyzes and reacts based on pre-defined rules. Think of it as a chef following a recipe. It’s the technology that gives you recommendations on Netflix or detects fraud on your credit card. In contrast, Generative AI, which is where a lot of the pioneering work is being explored, creates new content. This is where ChatGPT comes in?—?it is designed to help you be more productive. For example:

• You can give it text, and it can make it shorter, simplify it for a second grader or translate it into dozens of languages.
• With a simple prompt and some salient details, it can help you name your next program.
• It can turn a scenario into a country song or a rap.

But Generative AI can also do so much more.

• You can create and translate videos into many languages without a camera or crew (e.g., HeyGen).
• You can turn HR or training content into a podcast for your employees (e.g., Google’s NotebookLM).
• You can add Chatbots for common questions or needs on your website, reducing call or email volume.
• You can use it to help make a PowerPoint (e.g., Gamma).

And now nonprofits are getting into the game with custom applications. Google’s Generative AI Accelerator can help nonprofits build gen AI-powered solutions. Check out the 2024 cohort?—?it was incredibly impressive.

Can AI help with grant writing? What are the do’s and don’ts for our?writing?

This is a hot topic, because grant writing requires significant time and energy. Since many grants are connected to an existing boilerplate, Generative AI can be a great tool to increase your productivity. However, I have seen recent grants where AI was obviously used, with tell-tale signs like responses not matching the questions or the tone sounding robotic. And trust me?—?the review panel noticed and asked, “Can we trust this nonprofit?” So, be careful!

The following are helpful hints to make the most of Generative AI without going overboard:

Do:

• Use for first drafts (but always edit the response)
• Use to tweak language (for different audiences and purposes)
• Use to map out structure (how to section an article or chunk longer content)
• Use to change word count (how to synthesize writing into shorter lengths)

AI can also help with:

• Programs: Simplifying Theory of Change/logic model language, summarizing aggregate evaluation data
• HR: Writing job descriptions, interview questions and rating rubrics
• Fundraising: Creating thank you letter templates and tidying up grant writing
• Social Media: Editing social media post content and suggesting popular hashtags to boost visibility
• Communications: Writing the first draft of a press release and assisting with speech writing
• Governance: Drafting new policy or Terms of Reference templates

Don’t:

• Rely on AI exclusively for research purposes (it could be wrong!)
• Share confidential information (content pasted in an AI interface is put in the public domain unless you have a private account)
• Use without editing (ensure you tailor language to capture your individual or organizational personality)

Should we have a policy on use of AI and data management at our nonprofit?

Yes?—?we all need to stay ahead of the technology and ensure that donor and employee data as well as intellectual property is protected. Any policy you create should cover topics, such as:

• Ensuring your organization has a comprehensive AI Data Protection Policy
• Combating Plagiarism with AI
• Requiring accuracy through human oversight

So far, our favorite (free) template is available through Community IT Innovators.

Does AI really work? Can it produce something really?good?

This is always my favorite question?—?it usually comes from healthy skeptics. We recently saw a post from a nonprofit leader who asked ChatGPT to rework America Ferrara’s iconic ‘Barbie’ speech for the nonprofit sector. Here’s what it produced:

“It is literally impossible to run a nonprofit. You do so much good, make such a difference, and yet it never feels like it’s enough. Nonprofits are expected to be extraordinary, but somehow, they’re always told they’re doing it wrong.

You have to address huge societal problems, but you can’t spend too much money doing it. You’re supposed to pay your staff fair wages, but not so fair that donors think their money is being ‘wasted.’ You’re supposed to run lean operations, but not so lean that you can’t deliver programs. You have to spend every dollar wisely, but also innovate and grow.

You’re supposed to serve your clients wholeheartedly, but don’t forget to cater to funders, donors, and board members. You have to measure your impact constantly, but don’t spend too much money or time doing it. You have to market your cause, but not so much that it looks like you’re wasting resources.

You’re supposed to collaborate with other organizations, but not so much that you lose your unique identity or look redundant. You have to advocate for systemic change, but don’t get too political. You’re supposed to inspire trust, but if you admit to challenges or failures, it’s seen as weakness.

You can’t push back when a donor demands control over your programs, even though you’re the expert. You have to celebrate small wins, but also constantly remind people that the work is far from done.

You have to keep programs running, keep clients happy, keep funders engaged, and somehow, keep your team from burning out. And if you don’t, people assume it’s because you’re not trying hard enough or that your leadership isn’t strong enough.

It’s exhausting. Nonprofits are tying themselves into knots to meet these impossible, contradictory expectations, just so they can continue to do the work that everyone says is so necessary. And yet, instead of being celebrated, they’re criticized for what they’re not doing or could be doing better.

The truth is, no one can solve these problems alone, and no nonprofit can meet every demand. And if we keep expecting them to, we’re only setting them up to fail.”

Pretty impressive, right? AI and its cousin Generative AI are here to stay. It reminds me of when I gave up my cherished tape recorder (and cool mix tapes) for a CD player and then gave them both up for an iPod. And now I just use Pandora and my iPhone. I still have my CDs in storage, because I spent a lot of money on them, but times change. In fact, change is the only constant. As the quote from Buckminster Fuller above suggests, just think about something you need (e.g., a Powerpoint, a new logo, a training video) and then find an AI tool to help you. You will learn the power of AI by doing! At a minimum, I promise it will be fun and kickstart your process.

We’d love to hear how you are using AI/Generative AI to help supercharge your nonprofit’s mission. If you have follow-up questions or feedback, send it along. I promise we will NOT use a chatbot to answer them.