AI’s greatest risk isn’t technical. It’s human.
We build AI-capable humans in not-for-profits, NGOs, and community organisations, through behavioural awareness and practice.
Practical, self-guided AI resources, grounded in Australia and built on evidence. The communities most exposed to AI’s risks are served by the organisations least resourced to manage them. We exist to change that.
Human Behaviour Cards, available now
Where the risk actually sits
Nothing about this moment feels like a failure. That is the problem.
A staff member opens an AI-generated client summary. It sounds authoritative. It is well structured. Seventeen other clients are waiting. The summary is approved without being read carefully.
No policy was broken. No system malfunctioned. No one made a bad decision on purpose. But a judgement that used to happen did not happen, and nobody noticed it was gone.
This is where AI risk concentrates. Not in the model. Not in the platform. In the small, private decisions people make about when to trust AI output, when to check it, when to disclose it, and who is accountable for what it produces. Most of these decisions happen quietly, outside where any policy can reach.
Not For Humans works at that interface. The behaviours involved are not new. Deferring to confident voices, spreading responsibility until no one holds it, letting shortcuts become habits. People have done these things for as long as there have been people. AI is simply the environment where they now surface with new speed and new consequences.
Most governance frameworks stop at the outer ring. Not For Humans works at the centre.
Why human behaviour is the most important variableWho this is for
Built for the organisations that serve communities others overlook.
If your organisation exists to support people rather than generate profit, and AI is arriving in your work whether you invited it or not, this is for you.
Not-for-profit organisations
No governance function. No dedicated budget. AI arriving anyway.
Charities, community services, health and wellbeing providers, disability services, aged care, and housing. Adopting AI without the infrastructure that other sectors take for granted.
Non-governmental organisations
Operating on public trust, often in complex and resource-constrained environments.
NGOs working in advocacy, development, humanitarian response, and social change. Organisations where the gap between AI promise and governance reality lands hardest.
Community organisations
Relationships are not a feature of service delivery. They are how services work.
Grassroots groups, peak bodies, community centres, and place-based organisations. AI use that damages trust with community members does not just create a governance problem. It undermines the operating model itself.
On the record
Australia’s Senate Select Committee on Adopting AI found that the communities least equipped to respond to AI adoption will bear its heaviest costs. The organisations closest to those communities are the ones these resources were built for.
Senate Select Committee on AI →Where Not For Humans works
Most governance effort lands where the fewest problems start.
Technical systems and formal policies receive most of the attention. Human judgement and behaviour receive the least, yet that is where most problems begin. Not For Humans was built to work in the space everything else walks past.
80% of governance effort goes here
Technical and operational layers
Technical capability. Systems, models, platforms, automation.
Operational set-up. Policies, compliance, procurement.
Organisational systems and processes.
80% of problems start here
The human layer
How decisions are made under pressure.
What happens to judgement when AI output sounds certain.
Who owns an AI-assisted decision, and whether anyone checks.
When informal AI use becomes normal, and whether anyone notices.
Whether lived experience still gets heard when AI speaks first.
Not For Humans works here
Where to start
Start free. Go further when you’re ready.
You do not need a budget, a consultant, or a governance background to begin. Most organisations start with the free resources and go further only where the need is real.
Free resources
Understanding the AI Landscape
Plain English orientation. Read this first.
AI Authorship Statement
A one-page form. Who used AI, for what, and who is accountable.
AI Register Template
A standing record of every AI tool in use. Start it once, maintain it.
Not For Humans Behaviour Lab
Interactive scenario-based practice. No prior knowledge needed.
When you’re ready to go further
AI in Use Audit
Shows you what is happening in your organisation right now.
Leadership and AI Reflection Workbook
Each leader surfaces their own position privately, before group dynamics shape it.
AI Governance Policy Template and AI Governance Model
The documents and the thinking behind them.
AI Governance Action Plan
Turns intent into a committed, owned, dated plan.
Human Behaviour Cards
The full depth reference. 45 named risks, 45 better behaviours, and guided team activities for ongoing use.
AI and community impact
The downstream impacts are already measurable in Australian communities.
The people served by the for-purpose sector are among the most exposed to AI-driven change and the least represented in decisions about it. Entry-level work is contracting in AI-exposed roles. Service demand rises as disruption spreads. The organisations closest to those communities see it first.
Not For Humans tracks six areas where the human consequences of AI adoption are already visible, drawing on peer-reviewed research, sector data, and evidence from community practice.
Research note
Anthropic researchers tracking real hiring data against AI exposure found entry-level contraction in high-exposure roles is already measurable. Workers most exposed skew female, educated, and mid-career.
Read the research →
“Not For Humans exists because I’m genuinely worried about who will bear the human costs of rapid AI adoption. I know it won’t be distributed fairly.”
Adrian Hubble, Founder, Not For Humans
Not For Humans is a human behaviour business. The expertise is in the behaviour, not the technology. Every resource is low cost or free, self-guided by design, and built on the understanding that your organisation is already the expert in reaching your community. These resources exist to protect the judgement that makes that work possible.
We build AI-capable humans in not-for-profits, NGOs, and community organisations, through behavioural awareness and practice.