Nov 6, 2026
Ep 52: How Non-Profits Can Adopt AI Without Losing Trust with Michelle ‘MJ’ Johnston
Kate vanderVoort, Founder and CEO of the AI Success Lab, is joined by Michelle 'MJ' Johnston, founder of AI with Purpose, for the Skills in Practice series.
MJ spent 30 years in information technology, left it entirely to become a Benedictine monk and a jail chaplain, then came back when AI arrived. That combination now shapes a practice built around small not-for-profits - organisations with tight budgets, heavy staff turnover, and a reputation that rests entirely on trust.
Most AI conversations in the sector start with the tools. This one starts somewhere else.
Can an organisation whose whole value is trust adopt AI without spending that trust?
In This Episode We Explore
Starting With the Ethics, Not the Tools
One client came to MJ with a question before any decision about software: is AI ethical, and should we be using it at all. The answer took research, not reassurance.
- why the ethical questions have to be answered before a tool is chosen
- researching data centres, water use and electricity demand from both sides of the argument
- refusing to deny the real harms while showing what is changing
- the difference between being sceptical and being closed
- helping organisations advocate for better outcomes rather than opting out
Moving From Dabbling to a Working System
MJ describes the early stage most people recognise - fumbling with ChatGPT, chasing new tools, producing scattered results that never quite build on each other.
- why random prompting produces fun outputs and no momentum
- building the foundation documents that give AI real business knowledge
- using Notion as a command centre to track what was made, when and how
- the shift from generic output to output that already knows the organisation
- why the tracking matters as much as the making
Capacity, Not Replacement
A long-standing client wanted to be seen as a leader in their field but had neither the staff nor the time. MJ built the thing they could never get to.
- researching what sector leaders already published, then finding the gaps
- turning the organisation’s own documents into five detailed FAQs
- separate sections written for parents, clients and people working in the sector
- human review of every program detail before publication
- publishing with an open note that AI helped and humans directed and reviewed it
- why transparency protects reputation rather than risking it
Keeping Knowledge When People Leave
High turnover is normal in the sector, and every departure takes knowledge with it. MJ’s answer is procedure documents that record reasoning, not just steps.
- writing the human rationale into every standard operating procedure
- recording why a particular person reviews a particular thing
- two versions of each procedure, one for people and one for AI to run
- building in review dates so documents do not quietly go stale
- keeping an AI register so nobody inherits an automation they cannot find or switch off
Practical Takeaways
- Answer the ethical questions first, with real research, before choosing any tool
- Pay for a higher reasoning model rather than relying on the free tier
- Check the not-for-profit pricing - OpenAI, Anthropic, Notion and Canva all offer it
- Build foundation documents so AI knows the organisation before you ask it for anything
- Track what you build, when you built it and how, or you will lose it
- Write the why into every procedure, not just the how
- Use voice rather than typing when you want to think something through properly
- Review everything before it goes out, every time
- Ask whether the information you are protecting is already public before treating it as high risk
MJ’s Favourite AI Tools
- Claude and Claude Cowork - the command centre for foundation documents, procedures and brand voice
- Notion - tracking documents, review dates and governance research in one searchable place
- ChatGPT and Gemini - alternatives worth paying for to reach the higher reasoning models
- Nano Banana - quick image fixes such as removing power lines from a client photo
About MJ
MJ is the founder of AI with Purpose, helping small not-for-profits adopt AI without compromising their values. A background spanning 30 years in networks, systems and code, three and a half years as an Episcopal Benedictine monk and jail chaplain, and certification as a life coach, shapes an approach that treats ethics and safety as the starting point rather than an afterthought.
Work With MJ
MJ coaches business owners and not-for-profit teams through the 90 Day AI Skills Accelerator, the AI Success Lab program that takes you from scattered experiments to a working AI system in your organisation.
If the FAQ project, the foundation documents or the procedures built around human rationale sounded like what your organisation is missing, this is where that work gets done, with MJ alongside you.
Find out more and apply: https://aisuccesslab.com/90-day-ai-skills-accelerator
Connect with MJ
Organisation: AI with Purpose
Website: https://AIwithPurpose.ai
LinkedIn: https://www.linkedin.com/in/aiwithpurpose
Final Reflection
The pushback against AI in the not-for-profit sector is real, and much of it is reasonable. What this conversation shows is that scepticism and adoption are not opposites. The organisations doing this well are the ones asking the hardest questions first and then deciding, rather than deciding by default in either direction.
If your organisation’s greatest asset is trust, what would you need to know before you let AI anywhere near it?
No comments yet. Be the first to say something!