📨 AI for Social Impact Newsletter

👋 From the Editor

Hi! I’m Joanna. I’m on a mission to help folks in the social impact sector understand, experiment with, and responsibly adopt AI. We don’t have time to waste, but we also can’t get left behind.

Let’s move the sector forward together. 💫

🧠 EDUCATION

• Foundation Models: Foundation models are the backbone of tools like ChatGPT, Claude, and Gemini. Nividia explains, foundation models are AI neural networks trained on massive unlabeled data sets to handle a wide variety of jobs, from translating text to analyzing medical images. “AI neural networks”, modeled on the human brain, use artificial neurons that learn by strengthening or weakening their connections based on experience (“machine learning” or “deep learning”). This matters for us in the social impact space because these models are the “foundation” for most specialized tools, from grant writing assistants to program evaluation analytics. It’s also possible to retrain foundation models for specific use cases, but as always, the most important key is you, the human behind the output. 👥 💻

• Environmental Impact: It’s no surprise that one of the biggest fears around AI in the social impact space is the environmental impact. A University of Massachusetts study found that, in 2019, training a single AI model could emit as much carbon as five cars in their lifetimes. However, in 2025 Google released report showing that over 12 months, the median energy consumption and carbon footprint per Gemini Apps text prompt decreased by factors of 33x and 44x, respectively. It seems tools are gaining efficiency over time (AI training AI, anyone?)… and, we can also consider the flip side of using AI for environmental conservation efforts, like predicting climate patterns and tracking deforestation. So then the question is, how do we know if the costs outweigh the benefits? ⚖️

✨ INSPIRATION

• Foundation Models for Good: Foundation models are beginning to revolutionize how social impact organizations work. The Patrick J. McGovern Foundation’s Grant Guardian uses Claude to power its AI tool that streamlines the financial review process for both grantmakers and nonprofits. Serenas, a nonprofit dedicated to ending violence against women and girls in Brazil, uses ChatGPT to find funding sources and draft grant proposals in both English and Portuguese. Team4Tech also leverages ChatGPT to curate their EdTech resource hub serving over 800 organizations. ✨

• AI for Environmental Conservation: Earth Species Project is developing NatureLM-audio, the world's first large audio-language model for bioacoustics, training AI to decode animal communication from whales to crows (cool!). 🐳 Climate Policy Radar uses AI models to organize and search through thousands climate law and policy documents from countries all over the world, making climate legislation searchable for policymakers worldwide. 📄 And World Wildlife Fund in the Netherlands developed Forest Foresight, piloted in Borneo and Gabon, that uses AI to predict forest loss up to six months out with 80% accuracy. In Gabon, Forest Foresight helped rangers detect and stop an illegal gold mine, protecting about 74 acres of forest. 🌳

🚀 ACTIVATION

• Social Media Insights: Who has time to dig through social media metrics? Try this prompt to uncover actionable insights: "I'm analyzing social media performance for my [nonprofit type/mission]. Here's my data from the past [time period]: [paste your metrics like post reach, engagement rates, follower growth, top-performing content types, or upload a report]. Please help me identify 3 key insights. For each insight, include: the pattern you noticed, why this matters for nonprofit communications, and one specific action I can take this month to improve our social media impact. If you need any additional information to make your analysis stronger, please ask.” 📊

• Qualitative Data Analysis: AI can be your handy assistant in analyzing qualitative survey data. Sample prompt: "Please help me analyze [number] survey responses from [participant type] about [program/service]. Here are the responses to the question '[paste your question]': [paste 10-15 representative responses]. Please identify the top 3-5 recurring themes or patterns in these responses. For each theme, provide: a clear theme name, 2-3 supporting quotes from the responses, and what this suggests about [participant experience/program effectiveness/areas for improvement]. If you notice any concerning patterns or unexpected insights, highlight those too.” Side note: you can also anonymize your data and upload the survey results into your preferred LLM for direct analysis (and less copying and pasting!).  🔎

🤖 What I Asked AI This Week

Hey Claude, what is your spirit animal? 🐙

“That's a fun question! If I had to pick based on my observable behaviors and how I function, I'd probably say an octopus. Octopuses are incredibly adaptable - they can change their approach based on the situation they're in, much like how I adjust my communication style depending on what someone needs.

They're also known for being curious and exploratory, carefully examining things from different angles before responding. Plus, octopuses are surprisingly good at fitting into tight spaces and working within constraints, which is pretty much my daily existence as an AI with specific parameters!” 😆

💬 What Do You Think?

This newsletter is meant for you, so I’d love to hear what you think. 💌 Reply anytime!

👀 ICYMI

If you’re new here, welcome! You can check out the archive of past issues here.