AI Impact Hub · The AI Adoption Pathway
A living pathway that guides your organisation through adopting AI — application-first, self-paced, and alive as the landscape shifts. You don’t finish it. You walk it.
01 — The reframe
Most stall before the end. The few that finish are outdated the moment the models move.
Remembers where you are, and keeps producing value as the landscape shifts. There’s no finish line — you keep walking.
“A course is only as good as your ability to apply it to your work.”— Kyle Behrend, AI Impact Hub
Information isn’t the moat anymore. Getting people to actually adopt it — that’s the work. So we stopped selling information, and started selling facilitated change.
02 — Settled by the evidence
of one partner organisation’s team used AI weekly within a year — up from ~5%. A change arc, not a course.
of nonprofits have an AI policy — yet 90% are already experimenting. The gap is the front door.
of AI’s value is in work that didn’t exist before — not in optimising the old. Adoption unlocks it.
03 — The framework
The AI Adoption Pathway. One journey, walked once — then it becomes a loop that never ends. Every station leaves you a real artifact in your real work.
Who you are, size, mission, the tools already in use, how people use AI, leadership stance. Light — and it pre-fills everything downstream.
◆ Baseline “you are here” across all sevenThe builder interviews you and generates a real, plain-language AI policy — stored as structured, clause-level data.
◆ Your living AI policyRemoves the #1 barrier — “am I even allowed to use this?” Clause-level storage means the Radar can later point at the exact clause to revisit.
AI shines a light on a bigger problem — how your organisation handles data at all. Classify data green / amber / red, run a real data audit, assess every tool, then switch on the Radar.
◆ Full data audit + classification + tool assessments + live RadarAI doesn’t just add risk — it exposes the cracks already there. Adopting it forces the data questions most orgs have quietly avoided:
So a full data audit sits right alongside the AI work. Then the Radar keeps watch: external change → the exact clause in your policy or tool in your stack → a plain-English action. Almost no one maintains that for nonprofits.
Guided setup of the baseline every knowledge worker should have — chatbot, voice dictation, meeting notetaker, deep research, a first agent.
◆ A configured toolkit + who-uses-whatWith real setup checkpoints from the organisations we’ve worked with — “is it actually set up?” — so tools get used, not just installed.
The meta-skill under everything else: projects set up as contextually aware ecosystems, AI that interviews you one question at a time, and your way of doing things turned into frameworks and skills you own.
◆ Contextually aware projects + your first reusable frameworksEveryone rents the same models — the context and frameworks you build are the difference. Own the framework, rent the model: swap the engine underneath without starting over.
Curated best-in-class training routed by role — we point, we don’t re-make — plus the self-learning system: using AI to learn AI.
◆ Role learning routes + self-learning habitsAnthropic’s course, OpenAI Academy, the right YouTube — curated, not recreated. Deep use and deep understanding.
Install rituals — AI at the end of team meetings, a #ai channel, organic champions, room to say “it didn’t work.”
◆ A culture playbook + champion listMake use visible, safe, and a little fun. Adoption is a culture problem long before it’s a tooling problem.
Stand up your own maturity model — a general template, customised — and run a baseline pulse.
◆ Org maturity model + first pulse + movementNot a score you’re sold — a movement-over-time view that proves behaviour actually changed.
04 — The moat · change intelligence
The Policy & Security Radar watches what changes in the AI landscape and tells you exactly what to revisit — so your policy is never a PDF that rots. Hover a signal:
05 — Same journey, different support
The front door
Nearly every nonprofit knows they’re missing an AI policy. It’s the easiest yes — and the whole journey rides inside it.