Who among us hasn’t faced this problem? You’re happily working away at your fulfilling social impact job when someone asks you to look at some data. Maybe it’s reading some survey responses from your constituents, or maybe you have to find some figure in a report that prints out of some arcane CRM thing that someone set up decades ago. Either way, while you’d like to look at that data in a beautiful and meaningful way, you’re instead confronted with the WALL OF NUMBERS.

Ugh. You have to scroll through little cells, zoom in on bad PDFs, and jump through a million annoying hoops just to answer a simple question like “how many people are definitely coming to our event?”

Well if that sounds like your daily grind, we’ve got great news! Even your most basic AI tool can likely turn that mess of rows into something beautiful, interactive and, most importantly, customized to what you want to see in your data. Take a look at what Claude can do if you toss some 990s into it and ask for an interactive dashboard:

“BUT WAIT!”, you say. “I work with sensitive data! I’m not just throwing that into some tech company’s AI to get some pretty graphs!” Nor should you! So what do you do if you work with sensitive data? What if you’re not comfortable uploading a spreadsheet or PDF of your internal data to Claude or ChatGPT or Copilot, no matter how much their privacy settings assure you they absolutely promise not to let anything happen to it?

Try it out - and let us know how it goes!

That’s why we developed the FULLY PRIVATE, interactive data viewer recipe. All you have to do is use the recipe here: copy it into your AI tool of choice, and it will walk you through the process of making your *own*, custom dashboard that works completely on your computer in the browser, and never sends any data anywhere.

Just like that!

Custom data visualization with zero risk - what more could you ask for?

Keep your eyes on this spot for the release of the full Cookbook V2 later this month.

The cookbook was a perfect entry point — it provided them a chance to see the “art of the possible” by browsing prompts they could use to solve real nonprofit problems and it gave them a simple way to copy/paste their way into a solution.

It may not seem like August 2025 was that long ago, but it feels like an entire generation in terms of AI capabilities time. Since we launched the cookbook, the off-the-shelf capabilities of tools like ChatGPT and Claude have increased tremendously.

Aside from the obvious additions of whole new tools like Codex and Claude Code for writing code and Work/Cowork for running more ‘agentic’ tasks in businesses, even your standard ‘vanilla’ chats are capable of creating some truly impressive results with no coding expertise needed. For example, we now regularly watch our nonprofit cohort members go from their first prompt in ChatGPT to interactive dashboards of their data.

So we asked ourselves, “if the Prompt Cookbook was meant to be a starting place to show nonprofits what AI out-of-the-box could do for them without needing to know any super technical tricks, what would that look like today?”

That’s what inspired the Prompt Cookbook Volume 2. This edition is a set of new prompts that show off some of the new powers of these tools to act as different personas, like a Strategic Advisor that knows your company’s work, create repeatable processes like always creating materials in your brand colors, and even creating little standalone pieces of code that even non-technical people can create and run.

Try it out - and let us know how it goes!

The full V2 will be coming out shortly, but, while we put the finishing touches on, we wanted to share a few recipes we think are really unique, useful, and easy-to-run that most nonprofits probably don’t know about. The first one up is a Private Audio Transcriber. Access it here.

Many nonprofits we work with want the ease of transcribing conversations but understandably worry — who will see that transcript? Will sensitive populations’ data live in some company’s cloud? Could that get into the wrong hands?

If that sounds like you, this recipe is for you! This prompt will guide you through setting up your own audio transcription tool such that data never leaves your computer. It’s not a full featured replacement for Granola or Otter, but it’ll do in a pinch for the most sensitive of comments.

Play around with it here and let us know what you think at decodedfutures@technyc.org!

And stay tuned for more recipes to come in the new release. We can’t wait to see what you all build!

There’s a difference between using AI and using AI effectively.

We know first hand how tricky it can be to get the kind of output that actually saves time, strengthens programs, or helps a team work smarter. The secret is the prompt.

But here’s the thing: writing good prompts takes time. It takes pattern recognition, experimentation, and precision that’s hard to do between meetings or grant deadlines.

So, our entrepreneur-in-residence, Jake Porway, decided to bottle our approach to prompting.

The result: The Decoded Futures AI Cookbook, a collection of more than 360 tested prompts organized by real nonprofit challenges. Instead of every organization starting from scratch, the Cookbook offers a flexible foundation they can plug into, adapt, and make their own. These templates prevent nonprofits from reinventing the wheel, because while every nonprofit’s mission is unique, the challenges are often shared.

Thousands of organizations have already picked it up. Some are building workflows straight from its pages; others are remixing prompts to fit their mission.

How It Works (and Where It Started)

The AI Cookbook is organized around five types of problems we keep seeing in the field:

1. Create & Communicate: writing, storytelling, and translation

2. Automate the Admin": streamlining repetitive, time-heavy work

3. Answer & Assist: building helpers for staff, volunteers, and communities

4. Sort & Scan: organizing, evaluating, and prioritizing information

5. Learn & Decide: analyzing data and making evidence-based decisions

Each prompt recipe has three difficulty levels (low, medium, and high complexity) so a single nonprofit can start with a plug-and-play prompt and grow into a custom GPT or workflow.

The goal is to make AI useful and to turn “we should look into this” into “we already built that.”

We introduced the Cookbook at the Open AI Nonprofit AI Jam, where nearly 1,000 nonprofit leaders across the country built real AI tools in a single day. Decoded Futures designed the Jam Day training and facilitator guide that powered every city’s event, transforming a national experiment into a hands-on, repeatable model for learning with AI.

When we first tested this approach during the OpenAI Nonprofit AI Jam, the results spoke for themselves:

• 87% of attendees said the Jam was very helpful or excellent.

• 78% left with a working prototype.

• And 95% said they planned to keep applying AI to their work.

(Check out the full report on Jam Day from OpenAI.)

Those numbers confirmed what we’ve been seeing all year: once people get hands-on practice and the right examples, AI stops feeling abstract and becomes another tool in their problem-solving kit.

Why This Matters

Real adoption doesn’t come from webinars or white papers. It comes from trying something, breaking it, fixing it, and sharing what works.

Every prompt in the Cookbook started as a real problem: a grant deadline, a data backlog, a newsletter that had to go out by morning. Together, they map what’s actually happening across the social sector—the patterns, pain points, and breakthroughs that define this moment in AI.

By open-sourcing those learnings, we’re helping nonprofits everywhere skip the hardest part: the blank page. They can start where others left off, remix ideas for their own context, and spend more time delivering impact instead of deciphering technology.

It’s also a signal to funders and policymakers: the people most qualified to shape how AI serves communities are already doing it. They just need the tools, time, and trust to keep going.

Try it out - and let us know how it went in the comments below!

You can explore the AI Cookbook or read OpenAI’s Nonprofit Jam Report to see the national results.

Applications for our 4th Cohort of Decoded Futures is now open!

• Apply here to join the cohort if you are a NYC-based nonprofit leader

• Apply here to volunteer as a technologist

Applications are due December 21st - email decodedfutures@technyc.org with questions.

Last week, we shared the four ways AI leadership is emerging in the social sector.
This week, we’re zooming out to the wider view: how those same shifts fit into the global conversation.

The latest State of AI Ethics Report from the tracks how the field is evolving in real time, through the people building, questioning, and reshaping what AI can mean.

As part of it, Decoded Futures contributed How Nonprofits Are Using AI: What’s Working, What’s Not, and What They Need to Succeed, drawn from what we’ve seen firsthand in hundreds of organizations: the experiments, the dead ends, and the wins.

Read our contribution below, or explore the full report here.

Originally published in the Montreal AI Ethics Institute’s State of AI Ethics Report, Vol. 7, pp 150-152.

When Dave Hazan, Director of CEA Operations, at New York Sun Works trained an AI model to analyze program reports, he expected a small efficiency gain. Instead, the task that once consumed an entire workday now takes ten minutes. Across the social-impact field, similar experiments are transforming daily work. From museums generating personalized storybooks to libraries offering multilingual search, nonprofits are discovering that AI’s value lies in the quiet redesign of everyday tasks.

Decoded Futures, a program of the Tech:NYC Foundation, has trained over a thousand nonprofit leaders in hands-on, practical AI applications. Its training approach combines technical learning with human-centered design, helping participants translate new tools into meaningful practice.

Three key components make this learning stick:

1. In-person, exploration-based learning: Participants experiment with AI tools, connecting concepts directly to their own organizational challenges.

2. Tech expert support: Each cohort includes access to practitioners who can troubleshoot, recommend tools, and translate technical ideas into nonprofit contexts.

3. Workflow and problem mapping: Participants identify specific pain points first, then design clear workflows where AI can add measurable value.

Alongside these practices, every project is grounded in three ethical guidelines that ensure responsible and inclusive use:

1. Personally Identifiable Information: Free AI tools aren’t private; never share sensitive or personal data unless it’s through a secure platform.

2. Keep a Human in the Loop: Always review AI outputs before relying on them or sharing externally.

3. Customize for Your Community: Out-of-the-box models rarely reflect every culture or experience; nonprofits should adapt AI to serve their specific audiences.

Follow-up interviews reveal a clear pattern: AI has become a practical collaborator across fundraising, education, workforce development, and public service. 95% of alumni continued using AI after the program, building more than 200 new workflows. Their stories reveal what’s working, where friction remains, and what kind of support will determine whether this next wave of adoption advances equity or deepens divides.

What’s Working

• Everyday efficiency: AI is saving nonprofits time and mental bandwidth. At CUNY K16, Rachel Hutchins uses AI for charts and regression analyses that used to take hours. These changes shift effort away from repetitive work and toward mission-critical programs.

• Mission-aligned innovation: Organizations are using AI to enhance creativity and access. The Bronx Children’s Museum built a story-generation tool featuring Bronx kids, expanding literacy access. The Brooklyn Public Library launched a multilingual recommendation system across 60 branches. Both examples show AI applied with cultural sensitivity.

• From pilots to systems: Some nonprofits are embedding AI into core operations. At The Leadership Academy, Alexander Negron integrated AI into Salesforce and Monday.com to automate workflows and support staff learning. United Neighborhood Houses is building a dashboard linking 45 settlement houses, creating real-time visibility across 800,000 New Yorkers served.

• Internal capacity building: Adoption spreads fastest when someone inside champions it. At Berkeley College, Randy Gomez built a cross-department GPT to identify partnerships and trained colleagues. At CUNY K16, Rebecca Beeman trained 18 senior leaders and surveyed adoption rates to guide decisions. Such initiatives turn curiosity into institutional readiness.

What’s Not Working

• Institutional hesitation: At larger organizations, some alumni use AI daily but face a system-wide moratorium. Similar caution across large institutions slows innovation until policy catches up, forcing some staff to innovate quietly.

• Resource constraints: Smaller nonprofits often lack the engineering support necessary to maintain their tools. Innovators easily build prototypes but struggle to sustain progress without technical mentorship or funding.

• Weak measurement: Few organizations formally track AI outcomes. Most rely on anecdotes, which makes it harder to evaluate the results.

• Fragmented support: Leaders describe needing continuing education, troubleshooting spaces, and policy guidance. They want peers to learn with, not just one-off training sessions to attend.

What Nonprofits Need to Succeed

The next phase of AI adoption depends on ecosystem design, which includes shared infrastructure, peer learning, and steady investment.

• Community of practice: Networks or spaces to exchange examples. Peer learning transforms scattered pilots into a collective force for progress.

• Hands-on guidance: Continued office hours or access to experts who can help implement ideas in real time.

• Advanced training: Deep workshops on automation, workflow design, and ethics, especially those scaling from prototypes to systems.

• Organizational support: Writing AI policies, integrating tools, or connecting with funders. What nonprofits need most is patient capital for responsible scaling.

Funders can meet these needs by investing in communities of practice, applied learning grants, and shared ethical infrastructure.

The Path Forward

AI in civil society is evolving from pilot projects into everyday infrastructure. The leaders driving this change are proving that innovation rooted in mission can thrive without massive budgets. Even as funding is pulled back across the US, nonprofits are learning to scale impact and reshaping the social sector with AI. What they need now is sustained funding, technical mentorship, and communities that foster continuous learning.

That is how AI transforms the social sector, becoming a durable tool for equity and change.

Explore Montreal AI Ethics Institute’s State of AI Ethics Report, Vol. 7 here.

Decoded Futures is a program of the Tech:NYC Foundation that equips nonprofit leaders to use AI responsibly and creatively in service of their missions. Want to join the next cohort or learn more? Visit decodedfutures.nyc

Proudly funded by Robin Hood Learning + Technology Fund, Google.org, Salesforce, and Altman Foundation.

If you really want to understand how AI is reshaping the social sector, start with who’s learning to use it and how they’re bringing others with them.

This fall, Decoded Futures distributed its first round of grants to nonprofits across New York City (generously funded by Google.org and distributed by The New York Community Trust). These awards, a total of $1 million, mark a turning point: the moment AI capacity-building in the social sector moves from experimentation to investment. As we think about the sustainability and long-term impact of emerging technology in mission-driven organizations, one question keeps coming up: where should funders be looking next?

To answer that, we turned to the people who’ve been shaping this work from the start. Over the past year, Decoded Futures at Tech:NYC has worked shoulder-to-shoulder with more than a hundred NYC-based nonprofit leaders to test what AI can — and can’t — actually do for their organizations.

When we checked back in one year later with some of our earliest participants, some clear patterns had emerged. Namely, that AI adoption is spreading through people instead of mandates or corporate training programs, through leaders who are experimenting quietly, teaching their teams, and proving that responsible AI use is possible.

We call these patterns the 4 Archetypes of AI Leadership in the Social Sector: the Advocate, the Observer, the Experimenter, and the Builder.

Each one represents a different path for how technology becomes culture and a roadmap for what the social impact field will need next.

(Want to find out your AI Leadership style? Take our quiz here)

Where We Got the Data

These conclusions are based on a qualitative analysis of interviews with 22 alumni organizations of the Decoded Futures program, representing nonprofits across education, workforce development, health services, youth empowerment, and more.

Their stories, from internal training at CUNY K16 to new automation systems at NY Sun Works and AI-powered dashboards at SEO, reveal not just what’s possible with AI, but what conditions make responsible innovation stick.

From those interviews, four distinct leadership patterns emerged. Below we’ll unpack what each archetype reveals about how nonprofits are adopting AI, what they need to grow, and where funders can focus their support to make responsible innovation stick.

1. The Advocate: Making AI believable

Advocates are the nonprofit leaders who have gone from curious to catalytic. They’re the ones running internal lunch-and-learns, securing ChatGPT Plus accounts for their teams, or showing up at conferences to share what worked.

They tend to hold mid- to senior-level roles, which means they have both credibility and reach. Their power is social - not technical. They translate new technology into trusted practice, bridging the gap between program teams, IT, and leadership.

At CUNY K16, for example, Rebecca Beeman, Strategic Advisor to the University Dean of K16 Initiatives, ran her own post-program trainings with 18 senior leaders, surveyed adoption rates at one month and three months, and used the results to decide whether to renew licenses which built clear infrastructure.

What Advocates Need Next

Advocates need community. They want to keep momentum, but they need spaces and supports that sustain it:

• Peer communities for continued learning and support.

• Curated examples of real nonprofit use cases.

• Advanced workshops to deepen applied skills.

Where Funders Should Focus

Funders can multiply this group’s influence by supporting:

• Communities of practice: ongoing spaces for peer learning and resource exchange.

• Curated sector intelligence: newsletters or case repositories that highlight proven use cases.

• Intermediate training: applied, role-based workshops that go beyond introductory learning.

While funding peer-to-peer learning may not sound as exciting as building new tools, this is the foundation for how cultural adoption scales.

2. The Observer: Turning reflection into readiness

Observers are the cautious optimists, the AI users quietly piloting tools, watching how others use them, and thinking carefully about ethics and mission alignment. They’re often in leadership positions or large institutions where risk tolerance is low.

They balance curiosity with caution, and they bring a level of credibility that the sector desperately needs. Observers are the people others trust to keep things safe.

As Scott Hartl, President and CEO of EL Education, put it, Decoded Futures gave him a window into how peers were experimenting, a kind of informal field scan that helped him understand where to move next.

What Observers Need Next

Observers value structure and predictability over speed. They need:

• Structured spaces and recurring check-ins to benchmark timely progress.

• Ethical frameworks for responsible use.

• Low-risk spaces to test and learn safely.

Where Funders Should Focus

Funders can strengthen this group by investing in:

• Responsible AI frameworks: sector-wide guides for ethics, privacy, and risk navigation.

• Cross-sector benchmarking: shared metrics and case studies to compare adoption paths.

• Long-term learning infrastructure: recurring programs or alumni networks that normalize slow, steady adoption.

If Advocates spark momentum, Observers make it sustainable. They’re driving how AI adoption turns from novelty into widespread practice.

3. The Experimenter: Making AI real

Experimenters are the hands-on innovators, the nonprofit staff who learn by doing. They tinker with prompts, automate repetitive work, and test what’s useful.

They’re tech problem solvers who (usually) don’t know how to code. At NY Sun Works, for instance, Dave Hazan, Director of CEA Operations, trained an AI model to review program reports, cutting analysis time from a full day to ten minutes. Others, like Rachel Hutchins at CUNY K16, used AI to generate data visualizations that used to require hours in Excel.

They’re the ones who make AI visible as a quiet helper and integrated into everything in daily nonprofit work.

What Experimenters Need Next

Experimenters want to keep building, but they need lightweight scaffolding:

• Office hours and hands-on guidance for troubleshooting and workflow design.

• Toolkits and templates they can adapt without hiring engineers.

• Exposure to new use cases that spark fresh ideas for small-scale experiments.

Where Funders Should Focus

Funders should invest in:

• Applied training grants: small grants tied to experimentation with measurable outcomes.
  Technical mentorship: pairing nonprofits with technologists who can turn prototypes into usable tools.

• Mini innovation funds: flexible microgrants for testing, iteration, and adaptation.

The freedom to tinker like this will spur the next wave of nonprofit innovation.

4. The Builder: Turning experiments into systems

Builders are the organizational architects. They take what they learned as experimenters and advocates, and translate it into infrastructure. They’re the ones creating internal prompt libraries, embedding AI in reporting workflows, or coordinating across IT and program teams.

Builders lead through enablement instead of enforcement. They train their teams, standardize what works, and make AI part of the organizational rhythm.

Alexander Negron, Senior Director of Information Technology, at Leadership Academy, for example, used Decoded Futures insights to integrate AI directly into Salesforce and Monday.com, pairing it with ongoing staff training and office hours. His goal goes far beyond automation to making AI capacity part of the organization’s DNA.

What Builders Need Next

Builders are ready for scale but need structure. They’re seeking:

• Advanced build tracks that teach workflow automation and governance.

• Mentorship from technologists to help move prototypes into production.

• Peer cohorts to share lessons and avoid reinventing systems in silos.

Where Funders Should Focus

Funders can accelerate institutional change by supporting:

• Cross-functional implementation grants: funding for training, tech integration, and evaluation together.

• Capacity-building for responsible scaling: embedding ethics and documentation into growth.

• Shared infrastructure projects: platforms or tools that multiple nonprofits can use collectively.

Builders turn experiments into sustainable systems – they function as the connective tissue between innovation and operations.

What It All Adds Up To

These archetypes are stages of a growing movement: Advocates spread trust, Observers safeguard it, Experimenters test what’s possible, and Builders make it last.

Together, they form the early architecture of a distributed AI ecosystem in the social sector, one defined by trust, utility, and shared learning rather than hype.

For funders, this means shifting the focus from one-off innovation challenges to capacity ecosystems: communities of practice, technical mentorship, ethical infrastructure, and sustainable implementation funding.

The defining story of AI in the social sector will be about who learns to use AI tools and how they bring their communities along.

Want to find out your AI Leadership style and see how to drive implementation at your organization?: Take our quiz here

Decoded Futures is a program of the Tech:NYC Foundation that equips nonprofit leaders to use AI responsibly and creatively in service of their missions. Want to join the next cohort or learn more? Visit decodedfutures.nyc

Proudly funded by Robin Hood Learning + Technology Fund, Google.org, Salesforce, and Altman Foundation.

When our third cohort kicked off in late spring, it was clear an evolution had taken place. Our second cohort had just wrapped in March, so it had only been a matter of weeks, and yet the baseline of AI exposure among this new group had already risen. Instead of asking, “What’s a GPT?” they were asking, “Do I need a custom GPT?,” and were already experimenting with incorporating AI into their workflows.   

Our own baseline had shifted, too. The Decoded Futures team came into this cohort knowing how to build fluency and increase participants’ technical confidence; our challenge this time was more nebulous. On one hand, we had to convey the ambiguity inherent in AI-problem solving, namely, there’s no one best path to take. 

On the other hand, we needed the cohort to understand how to think computationally, because that’s what will lead you to identify what your problem actually is, in the first place. 

Here’s what we learned after our third Decoded Futures cohort: 

Embrace ambiguity.

Over the last year, we’ve seen three distinct patterns emerge in relation to ease of AI learning. As with any skill, people will take to it at different speeds:

1. Group 1: Fully embraces the ambiguity of using AI and is excited about the challenge. 

2. Group 2: Experiences some growing pains, but eventually digs in and finds solutions. 

3. Group 3: Wrestles with the inherent ambiguity of AI-problem solving and consistently feels a lot of discomfort in the ambiguity. (We get it!)

Interestingly, these groups don’t map to job titles, demographics, or even prior AI usage. The group that took to AI-problem solving the fastest – Group 1 – didn’t have any special training or characteristics. What made the difference for them was comfort sitting in ambiguity. Think of learning to float: it comes much easier if you’re relaxed and lean into it. 

Computational thinking is essential. 

This cohort arrived raring to build; some members had already made a custom GPT and others were even generating functional software through “vibe coding.” Technologically, the results were more advanced than what we’d seen in the past. But it was also clear that having exposure to AI tools (trying out the latest wrapper models or LLM) didn’t necessarily correlate to understanding how to use those tools most effectively. 

And you won’t know what tool makes the most sense for a given problem if you can’t really clearly identify what that problem is (it may not be what you originally thought!). 

Enter: computational thinking. At its most basic, computational thinking means breaking down a problem systematically. In our program, we teach members to map out their existing workflows with as much detail as possible. 

Let’s say you’re applying for a grant. Before you even open a document to begin drafting, you need to identify funding opportunities, review guidelines, gather data, and contact stakeholders. 

Looking at each discrete step in a process reveals the places where AI might be able to streamline or support. Alternatively, you could discover that AI isn’t necessary for a particular blocker at all, and it’s human interaction – having a conversation with a coworker, let’s say – that’s required. 

That distinction is important! The goal of our work isn’t to create AI evangelists, or use AI for the sake of using AI; we want to equip nonprofit leaders with a way of thinking that will empower them to work through almost any problem. 

The power of structured exploration. 

The feedback we received most often – and which won’t be surprising to anyone who’s ever struggled to sit down and tackle an extracurricular project, no matter how interesting – revolved around the value of these three program aspects: 

1. Dedicated, structured, recurring time, with encouragement to explore. 

2. Pairing with volunteer technologists from the very beginning. 

3. In-person peer support for real-time problem solving and brainstorming. 

This structured environment created momentum through shared discovery while ensuring no one got left behind.

Success with AI isn't about finding the "best" tool or following a perfect playbook. It's about developing comfort with ambiguity, thinking systematically about problems, and having support to experiment and fail.

Finally, as this cohort showcased so well at their recent Demo Day, you build and refine and iterate and fail and fail some more, and “that whole process of trying and failing, then failing better, is what we mean at Decoded Futures when we talk about “success.”

What’s New?

• Eng(INE) offers nonprofits free, dedicated AI engineering teams to build cutting-edge technology advancing their organization’s social and economic opportunity goals.

  • Applications for their next cohort are open until July 30! Check out their website to apply and learn more: nonprofitengine.org. 

• Bethany Crystal is running a virtual Summer AI School August 11-15. Join her to build a tool - or a business!

• We’re cooking up something big with OpenAI — keep your eyes peeled for an update on July 17! Until then, you can reach us at decodedfutures@technyc.org for any questions, comments, or suggestions.

That’s the vision which Jake Porway, Decoded Future’s entrepreneur-in-residence, laid out in his keynote address at the Robin Hood X Opportunity AI summit earlier this month.

Speaking to a packed crowd, Jake argued that "The social sector is the only sector that can and must solve AI alignment. If you put the nonprofits at the center, then we can define how technology works for us. We can use the same algorithms that Netflix is using to recommend movies to recommend social services. We can go from wondering, ‘Oh my God, how is my life going to change with agents’ and realize, you know what, we actually are the agents of change.”

Want to learn more about how we're putting nonprofits forward as agents of change? Make sure to subscribe to our substack, and reach out decodedfutures@technyc.org if you're a nonprofit or foundation interested in advancing AI for social good.

And don’t miss Nonprofit ENG(INE)’s virtual Demo Day on June 25, for 12-1pm ET, featuring product builds from Planned Parenthood, Mobile Pathways, and more. Afterwards, if you’re inspired to join their next cohort, register for their info session here (June 26, 5pm ET).

Share

Leave a comment

This is a guest post from Jake Porway, our entrepreneur-in-residence and a long-time AI-for-Good expert in the space.

Introduction

A wise person once said that a problem well-defined is half solved, and this couldn’t be more true than when it comes to building technology. I’ve spent fifteen years working on applying tech for good, and regardless of the tool or trend, the most common challenge I’ve seen hasn’t changed: understanding what, exactly, the new tech can be used for.

Many nonprofits are being nudged by their boards and funders to “use AI” without understanding how to apply it strategically across their organizations. This is why we spend so much time at Decoded Futures helping our learning cohorts define what an AI-shaped problem actually looks like.

We’ve worked with over 80 nonprofits on custom AI solutions, and in that process, we've identified five common use cases where Large Language Models (LLMs: tools like ChatGPT and Claude) or off-the-shelf tools can make an immediate difference.

Below, we’ll walk through each use case, identifying how to use AI in ways that truly maximize your impact.

1. Create and Communicate

“I need to write or create something to share with others.”

Whether you need to draft a grant proposal or write a bedtime story, this is a scenario where large language models (LLMs) like ChatGPT and Claude really shine. Though you aren’t limited to writing alone; from donor communications to educational resources, LLMs can support nearly every type of content you’d want to share with stakeholders.

Take The Reading Institute: During one of our AI learning cohorts, the Institute trained a custom GPT to generate “decodable books,” a highly specialized phonics tool to support early readers. Creating these books would usually take months and cost upwards of $3,000. With the help of AI, the Institute produced high-quality versions in a fraction of the time and cost, getting critical resources into the hands of students much faster.

Tools for Content Creation

• Gamma – for training materials and presentations

• Canva (with AI features) – for visual materials

• ChatGPT, Claude, or Gemini – for written content

Prompting Tips

When using general tools like ChatGPT or Claude, we recommend a two-part prompt strategy:

1. Set the tone. Describe your organization’s voice clearly, or upload examples of existing materials so the AI can match your tone and style.

2. Structure the output. If you need content in a specific format (like sections of a grant application), break down your instructions by section. Using Markdown or JSON formatting can also help ensure consistent results.

2. Automate the Admin

“I keep doing the same tasks over and over, and it’s taking time away from more important work.”

Repetitive, language-based tasks with clear rules are ideal for automation; think of reviewing timesheets or copying data between systems. At NY Sunworks, staff had to manually check volunteer reports for an entire school system and email individuals whenever corrections were needed; a process that was tedious, time consuming, and required little to no critical thinking. By training a custom GPT to review the reports, flag issues, and summarize fixes, NY Sunworks saved their staff hours of work, freeing them to spend more time on meaningful, mission-driven projects.

Tools for Administrative Automation

• Zapier for connecting systems and automating workflows

• Make (formerly Integromat) for more complex sequences

Prompting Tips

To get started:

1. Clarify the rules. Explain what a human reviewer checks for and how they make decisions.

2. Design the workflow. Set up a system that sends data to the AI and uses the output to take action, such as flagging errors or sending updates. You may need light coding help or support from ChatGPT to get started.

3. Answer and Assist

“I get lots of questions and can’t keep up with answering them all.”

If you’ve ever run a volunteer program, you’ve likely seen that certain questions get asked over and over again. Rather than relying on a stretched-thin staff member to respond, AI-powered question-answering systems, or RAG (Retrieval-Augmented Generation), can understand the question being asked and deliver a helpful, tailored response based on whatever internal information and resources you’ve supplied it with.

For example, Teach For America joined Decoded Futures to build a support tool for new teachers. With thousands of teachers onboarding each year, the TFA team needed a way to quickly answer common questions about training, school policies, and logistics. By uploading their materials to Google’s NotebookLM, they created a chatbot that could respond to teacher questions instantly, no coding required.

Tools for Q&A Automation

• NotebookLM for internal knowledge-based answers

• Chatbase for no-code chatbot setup

• Zendesk Answer Bot for support integration

Prompting Tips

1. Upload core materials like policies, FAQs, and guides; excluding private data.

2. Share examples of strong and weak responses to train the model.

3. Map question types to structured responses or decision flows.

4. Sort and Scan

"I spend too much time reviewing information that follows a pattern.”

If you’ve ever had to review candidates or applications through the lens of a consistent rubric, you know that this is essential work, but repetitive and time-consuming. An ideal AI-shaped problem! AI can score and prioritize hundreds of items based on criteria you’ve supplied it with, surfacing top contenders and allowing you to more easily make final decisions.

Project Basta used to manually scan job boards to identify roles that would be good matches to their recent graduates. This meant a staff member had to comb through a given job site, parse what skills each job actually required (not all “data analyst” positions are the same, for instance) then match that job to a particular graduate. Through Decoded Futures, their team built a custom GPT that was able to do this same process with 95% accuracy. Their use case was so successful that they’re now scaling in a major way, creating a full-fledged AI system for their career coaches.

Tools for Filtering and Prioritization

• Airtable AI for structured scoring

• Monday.com AI for team-based evaluation

• Gradescope for academic rubrics

Prompting Tips

• Define your rubric clearly, including scoring scale and weighting.

• Provide examples of what good, average, and weak entries look like.

• Ask the AI to rate how confident it is in each assessment

• Keep human review in the loop, especially for high-stakes decisions.

5. Learn and Decide

“I want to learn something new or make smarter decisions.”

Whether you're exploring a new strategy, analyzing survey data, or researching policy, AI can help gather insights, connect dots you might miss, and even act as a brainstorming partner, prompting you to think through complex issues.

CUNY’s K16 Initiatives needed to analyze a huge amount of student survey data; a perfect task for ChatGPT. They discovered that ChatGPT was able to transform qualitative answers into quantitative data. Their lead data analyst then used GPT to build a custom dashboard that grouped students based on response patterns. While other statistical software could also do this, it would have required far more time and technical effort, with no guarantee of outcome. Thanks to AI, the analyst had an easy, assured solution that didn’t rely on any specialized skills or niche programs.

Tools for Research and Analysis

Research

• Perplexity for web-based research with sources

• Wonder for fast, affordable research summaries

• NotebookLM to analyze documents and reports

Data Analysis

• Julius.ai for accessible data insights

• ChatGPT, Claude, and Gemini for smart, reliable analysis

Prompting Tips

For Research

• Ask AI to explore different sides of a question

• Request pros and cons of key approaches

• Prompt it to challenge its own assumptions

For Data

• Upload data to tools with robust analysis capabilities like Claude or ChatGPT

• Ask for both descriptive insights (what happened) and predictive insights (what might happen next)

• Request visualizations to help communicate findings to stakeholders

You Can Start Now

You likely can relate to at least one of these use cases. The hope is that you now also feel more empowered to begin testing AI solutions. Start small, with low-stakes projects. As you build confidence, scale up to bigger ones. Don’t expect to become an AI expert overnight! The goal isn't to use AI just for the sake of it, but to free up time and energy that can be better directed towards your actual mission.

The question isn’t if AI will change nonprofit work. It’s how soon an organization wants to lead that change. But you don’t have to go it alone! If you’d like to learn alongside other nonprofits as we all step into this new world, join our next Decoded Futures cohort. Happy hacking!

Tools, Tips, & Takeaways

Apply
Nonprofit ENG(INE) equips nonprofits with dedicated AI engineering teams to build cutting-edge technology that advances social and economic opportunity goals. Register for the info session here (6/26, 5pm ET). Applications open on 6/16 and close on 6/30; apply here.

Decoded Futures Cohort: Work alongside nonprofit leaders with support from technologist volunteers to build foundational AI knowledge, integrate tools into your workflow, and discover transformative possibilities for your organization.

Attend

Decoded Futures+Tech:NYC Social Impact AI Showcase: This is a rare opportunity to witness what human-centered, purpose-driven AI looks like in action. You’ll hear directly from the people behind the work, and what it really takes to move from experimentation to long-term adoption. (6/6, 2-4pm ET, New York, in person)

IBM AI for Impact: Learn to incorporate ethical practices, explore real-world fundraising applications, and develop actionable steps to integrate AI into your nonprofit’s work. (6/11, 830am-4pm ET, New York; in person)

Nonprofit ENG(INE)'s Demo Day: Interested in applying to our summer 2025 cohort? Learn from current participants Planned Parenthood, Upsolve, Recidiviz, and Mobile Pathways, as they demo their latest product builds. (6/25, 12-1pm ET, Virtual)

We’re standing at one of those rare, tectonic shifts in society.

AI isn’t just another gadget. It’s a cultural revolution, on the scale of the industrial era and it is changing how we work, learn, and lead. It is redefining who gets to shape the future.

But here’s the problem: nearly all that momentum is taking place in the private sector. It’s being used to optimize ads, scale profits, and build the next viral app. Meanwhile, the people solving society’s hardest problems, often with the fewest resources, are barely in the room.

That’s not just a missed opportunity. It’s a systemic failure.

Decoded Futures is here to change that.

We’re NYC’s AI-for-social-impact lab: part sandbox, part bootcamp, part matchmaking service between nonprofit leaders and technologists. Our mission is to ensure nonprofits and community organizations don’t just keep up with AI. They help lead it.

(Learn about our program here.)

We Know This Works Because We’re Doing It.

We’ve already helped over 80 nonprofit organizations go from AI-curious to AI-capable. That looks like:

• A $1.3M grant application won by an alum powered by AI learning through our program.

• Phonics books that once cost $3,000 a set will be free and accessible to every child in New York City thanks to a custom GPT built to democratize literacy.

• An alum equipped (and charged) to lead their 600 person organization in leveraging AI to increase their impact across NYC.

(See more Decoded Futures case studies.)

That is the power of applied AI. Not just dashboards, but real outcomes for real people. We’ve tested toolkits, run cohorts, and published insights in partnership with some of New York’s most mission-driven organizations.

Our team has spent the better part of 15 years in the trenches of the tech-for-good movement: building programs at major foundations, leading civic tech initiatives, and supporting community organizations through digital transformation. We've seen what works and what doesn't, and we know that building capabilities is a moving target when you're at the frontier of new technology.

(Check out our AI Adoption Report from our pilot cohort to see what we’ve learned so far.)

We’re Here to Build Bridges.

We know that learning together is one of the most critical needs in this early phase of AI. With support from our funders, the Tech:NYC ecosystem, and a growing network of technical mentors, Decoded Futures is creating a rare space. It is a place where the people solving society’s toughest problems can work alongside those building its most powerful tools.

And this Substack is your invitation into the room. We’ve created this space to share what we're learning and, more importantly, to invite all of you to share what you're seeing, too.

What You’ll Get Here

This is your front-row seat to the intersection of AI and social impact. You can expect:

• Stories from the field: what’s working, what’s not, and what’s next

• Lessons from our cohorts and pilots

• Frameworks and tools you can use

• Honest reflections on building with (and sometimes around) technology

If you care about social impact, innovation, and what it means to lead in a time of radical change, you’re in the right place.

Stick With Us.

Decoded Futures is where AI meets impact and we are just getting started.

Subscribe. Share. Or join us:

• Technologist: Volunteer your skills to support nonprofits with AI

• Nonprofits: Partner with us to explore AI

• Reach out to fund this work to ensure the AI revolution benefits everyone

- The DF Team