Case Study - AI adoption strategy for humanitarian and development organizations

#The challenge

Over the past year we have worked with several humanitarian and development organizations navigating AI adoption. Across each we saw the same starting point: staff were already using ChatGPT, Claude, or Copilot informally, but there was no shared understanding of what was safe, what was useful, or how AI fit the wider digital strategy. Many were paying for Microsoft 365 licenses with AI capabilities built in but weren't using them.

For most of these organizations, adopting AI isn't optional. Their teams are already stretched, managing complex programs with limited capacity, and most are already using AI in some form without getting the full value from it. AI can meaningfully reduce that burden but only if it's introduced with clear boundaries around data and realistic expectations about what it can and can't do well.

#Our approach

We meet each organization where they are and build from there.

#1. Understanding where things stand

We review how teams work: the tools they use, how data flows between them, where things break down, and what's going unused. We talk to staff across functions and place the organization on our AI Maturity Ladder (Crawl, Walk, Run) so everyone shares a language for readiness.

Most organizations sit in early-to-mid Crawl: some experience with AI tools, but no governance, shared practice, or clear path forward.

#2. Defining the path forward

From the review we build a staged plan sequenced around the organization's capacity, existing licenses, and priority workflows. This includes identifying which processes benefit most from AI or automation, setting data boundaries through a usage policy, and recommending tools grounded in what the organization already has.

We evaluate tools against real use cases rather than vendor marketing. Some of the use cases we have helped organizations develop:

• Building and maintaining integrations. Tools like Power Automate can be technical and time-consuming to configure. AI helps staff develop and troubleshoot automations without deep technical skills for every change.
• Developing on-brand design systems. Using AI design tools like Claude Design, we help organizations turn their brand book into an interactive system for producing on-brand slide decks, reports, and visuals without a designer in the loop every time.
• Prototyping ideas before building them. Non-technical staff use AI to mock up what they want, making it easier to communicate requirements to developers or vendors.
• Managing vendor delivery. AI helps teams groom project boards, identify completed tasks, flag gaps, and keep vendor accountability clear.

#3. Building capability that lasts

AI moves too fast for one-time training to be enough. What matters more is changing how people work: building the habit of experimenting with prompts to solve everyday challenges, learning what AI is good at (identifying patterns, summarizing, drafting) and what it's not (judgment, sensitive decisions, anything requiring institutional memory it doesn't have).

We help teams develop this mindset and the discipline around it. When someone finds a prompt or workflow that works, we help them document it so the AI can reference it in future work and others can build on it. Over time this creates a shared library of proven approaches that grows with the organization.

#What we have learned

These engagements have shaped principles we bring to every new partnership:

• Meet organizations where they are. The maturity ladder gives us a shared language for readiness.
• Start with data, not features. What must never enter an AI tool shapes everything else.
• Leverage what you already pay for. Most organizations are underusing AI capabilities already in their Microsoft 365 licenses.
• Build habits before systems. Weekly comfort with AI matters more than rushing to connect tools to organizational data.
• Plan for sustainability. Anything we introduce has to survive without us.

We share what we learn publicly. Our AI Maturity Ladder for NGOs blog post distills the adoption framework that underpins this work. Our guide to AI training for nonprofits reviews the courses we recommend. Future posts will cover building prompt libraries and skills, configuring AI tools for data protection, and designing AI governance for distributed teams.

#Technologies used

At Hikaya, we help NGOs and nonprofits adopt AI at a pace that fits, with honest reviews of where you stand, practical policies, tool evaluations, and staged plans that respect the sensitivity of your data and the capacity of your team.

If your teams are already experimenting with AI but you don't yet have a clear position on what's safe and what's worthwhile, start a conversation with us, or explore how we work.