AI SKILLS ATLAS
A Field Guide for Nonprofit AI Teams

What your peers actually do with AI — and how they learned to do it.

Synthesized from three April peer-learning sessions across Programs, Data, Leadership, and Fundraising tracks. This is a working resource: explore real use cases, find tools that fit your task, and pick a learning path that matches how you actually build skills.

27
Use cases
15
Tools peers use
3
Learning paths
8
Prompt techniques
01 / Use Cases

Real tasks, real peers, real workflows.

Filter by function. Click any card to see how a peer is actually using AI for that task — and what tool they reach for.

Filter →
02 / Tools

No single tool wins. Match the tool to the task.

Peers consistently named knowing which AI to use as its own skill. Here's the running cheat sheet from the sessions.

03 / Learning

Three paths to build your skills.

Peers learn AI in different ways. Pick the lane that matches how you naturally build new skills — or run all three in parallel.

01
The Tinkerer

By far the most common approach in the cohort. Pick a real task, try a tool, refine the prompt, repeat. No course required.

  • Choose one repetitive task this week
  • Try it in two different tools and compare
  • Save prompts that worked in a personal playbook
  • Ask the AI to suggest a better version of your prompt
  • Cross-check output with a coworker before sending
02
The Community Learner

For people who learn faster from watching, asking, and discussing with others. Peer sessions, coworkers, and online communities.

  • These peer learning sessions
  • Internal team lunch-and-learns
  • Reddit for prompt help and tool tips
  • Form an internal AI committee or working group
  • Talk to people outside your sector about how they use AI
03
The Curriculum Builder

For people who want structured input: courses, newsletters, video libraries, expert voices to follow over time.

04 / Prompts

Tactical phrases that changed peers' results.

Click any technique to see the prompt and why it works. Copy and adapt to your own work.

05 / Cautions

What to watch for.

Concerns peers raised in every session. Worth naming before you scale AI in your work.

Hallucinations & invented sources

Models make up citations confidently. Verify every factual claim, link, and statistic before forwarding. AI is a draft partner, not a fact-checker.

Privacy & PII

Never paste client names, donor data, medical info, or other identifiable details into consumer AI tools without organizational approval. Redact first.

The "AI tell"

Readers spot AI writing easily — em dashes, "delve," generic structure. Always edit for personal voice and specific detail before sending.

Format friction

Output rarely transfers cleanly between Word, Google Docs, Canva, and Excel. Build cleanup time into your estimates — it's real.

Tool sprawl

Audit your AI subscriptions periodically. Redundant tools waste budget and fragment knowledge. Pick a small number that work, get good at them.

Human services & connection

For mission-driven work, ask: what part of your job would you outsource? The honest answer reveals where AI fits and where the human matters most.