For organisations
AI for Orgs, Policy & Strategy
Two ways we help your organisation use AI well — practical team training, and the strategy, policy and governance that keeps you on the right side of the law.
Why this matters now
AI is already inside your organisation. Is anyone steering it?
Staff are using AI to draft, summarise and research, often on personal accounts and without guidance. The tools are moving faster than the policies, the training and the board's oversight. That gap is where both the risk and the opportunity sit.
Sources: Charity Digital Skills Report 2026; Schöbi, "Fast-tracking the use of AI in everyday philanthropy", in The Routledge Handbook of Artificial Intelligence and Philanthropy (2024).
From the research
“The underuse of this technology within the non‑profit sector may in itself be a form of misuse.”
Closing reflections, The Routledge Handbook of Artificial Intelligence and Philanthropy (eds. Ugazio and Maricic, 2024), from the University of Geneva's AI and Philanthropy project.
Staying away from AI is not the safe option. Organisations that hold back fall behind, while their staff use it anyway, without guidance. The answer is to use it well: with the right tools, clear rules and people who know what they are doing.
What we see
Four risks nobody notices until something goes wrong
Each one is preventable, and none of them is a reason to stay away from AI.
Data in the wrong account
Donor lists, case notes and board papers pasted into free, personal AI tools, where what you share may be used to train the model.
Confident, wrong answers
AI rarely says when it is unsure. A misquoted figure or an invented source can reach a funder, a regulator or the press.
Stories that cost donor trust
Research cited in the Routledge Handbook found stories labelled "AI-generated" were rated less engaging, even when a person had written them. How you use AI in fundraising, and what you disclose, matters.
Fraud at machine speed
Cloned voices, invented donors and automated "donor bots" gaming matched-giving campaigns. Old fraud, made cheap and convincing.
Why work with me
Research-led, built for real organisations
My work draws on my research at Oxford Saïd Business School on AI and systems leadership in philanthropy, the University of Geneva's international conference on AI and philanthropy, and more than 25 years building and funding start-ups, foundations and organisations across philanthropy, anti-money-laundering and counter-terror financing, fintech and investment.
“How can philanthropy leverage AI within systems leadership to effectively address complex global challenges, ensuring ethical application and preserving the human connection?”The research question behind my Oxford paper, The Resilient Leader (2024)
My research argues that the real challenge is not technical. It is leadership: using AI in a way that keeps the trust, transparency and human dignity philanthropy depends on. It comes down to six questions every leadership team should be able to answer.
Trust and authenticity
How do we use AI without eroding the trust of donors and beneficiaries, in an age of deepfakes?
Privacy and data security
Where is the line between using data for good and protecting people's rights?
Bias and fairness
How do we stop AI reproducing old biases in who receives resources and opportunity?
Access and inclusion
Will AI widen the gap for smaller organisations and the communities we serve, or close it?
Governance and accountability
Who is accountable when an AI-assisted decision has real consequences?
Human judgement
How do we make sure AI supports the judgement philanthropy depends on, and never replaces it?
I help teams get the hours back, and help boards and trustees answer these six questions with confidence. Two ways to start:
AI training for your team
Practical, hands-on AI training
Every member of staff who attends should leave able to use AI on a real task the same week. That is the aim. Suitable for colleagues at very different starting points — from those already experimenting through to those who have not yet begun.
The training is built on daily practice in real charity and foundation work, not theory or hype. Sessions are hands-on working sessions — people try the tools on real examples, not slides about features.
How we work
Three principles
Hands-on, not a lecture
Sessions are working sessions. People try the tools on real examples, not slides about features.
Built around your work
Examples come from your actual departments — fundraising, communications, events, administration, operations.
Honest about limits
We cover what AI is genuinely good at, where it falls short, and when to check its work. A guiding rule: if you wouldn't trust a sharp junior colleague to do it unsupervised, don't trust AI with it either.
What the programme includes
A programme you can shape
Components can be taken individually or combined. Most organisations start with the introductory session and build from there.
Introductory session for all staff
90 minutes · live on Zoom · recording shared afterwards
A practical, hands-on introduction to everyday AI use, suitable for everyone regardless of starting point. It demystifies AI, shows the small number of habits that make the biggest difference, and includes a live demonstration of a real task moving from a weak result to a genuinely useful one. Beginners leave confident enough to start; existing users pick up sharper technique.
Practical examples by department
Tailored to fundraising, communications, events, administration and operations
So the session lands for every team, it draws on examples tailored to each function — donor and grant communications for fundraising, press releases and content calendars for comms, logistics checklists for events, minutes and templates for administration, and SOPs and board reports for operations.
Responsible use, data protection and GDPR
60–90 minutes · can run within the intro session or standalone
What information should and should not go into AI tools, how to spot and check confident-but-wrong answers, and simple guardrails the team can adopt. Covers data minimisation, lawful basis and consent, appropriate tool settings, and sensible retention — practical GDPR-aligned guidance, not legal advice.
Responsible storytelling and fundraising with AI
90 minutes · fundraising and communications teams
How to use AI in appeals, case studies and donor communications without losing donor trust. Drawing on the Routledge Handbook of AI and Philanthropy: what research shows about how donors respond to AI-generated stories, bias in AI images and text, when and how to disclose AI use, and an ethical check before any campaign goes out. Includes a workshop rewriting one of your real appeals.
Money, fraud and AI
90 minutes · finance teams, senior staff and trustees
How AI is making old fraud cheap and convincing: cloned voices, invented donors, AI-written appeals and automated abuse of matched-giving campaigns. Grounded in anti-money-laundering and counter-terror financing practice, with simple checks that cost nothing, and how AI can help small finance teams spot unusual patterns.
Team programme (mirrors Arakura 101)
4 weekly sessions · up to 12 staff · private cohort
For teams who want to go further. The same four-week curriculum — Foundations, Context, Workflows, Compounding — run privately for your organisation. Staff finish with a personal toolkit of prompts, templates and workflows built around their actual roles, plus recordings of every session. Run across a team it builds shared habits and a common language, so departments work in alignment rather than each person figuring AI out alone.
Optional: advanced strand and Claude Code
Add-on for technically-minded staff · operations and data teams
For a small group who want to go deeper: deeper reusable workflows and an introduction to Claude Code for automating practical tasks such as data clean-up, recurring reports, and light website work. Genuinely useful but not for everyone — offered as an add-on for those who would benefit rather than core training for all staff.
Get started
Talk to me about training for your team
Tell me about your organisation and what you need. I'll reply personally, and all formats can be tailored to your budget.
AI strategy & policy
Strategy and policy, made clear.
AI is already inside your organisation, whether or not the board has decided it is. The rules are arriving quickly, in the UK and the EU. I help charities and nonprofits understand what is coming, get ready, and use AI with confidence, in line with the law.
Practical guidance to help you decide well. This is not legal advice. Current as of July 2026.
The governance gap
The gap between using AI and governing it
Most organisations have already started using AI. Far fewer have put any policy or oversight around it. That gap is where the risk sits, and it is exactly what trustees are responsible for.
Figures are indicative and drawn from 2026 charity sector reports. Trustees remain responsible for AI use under their existing duties in the Charities Act 2011, whether or not the board is aware it is happening. The 2025 Charity Governance Code refresh now expects charities to have a technology and AI policy.
The road ahead
What is coming, and when
The AI rulebook is being written in phases. You do not need to track all of it. You do need to know which dates touch your organisation. Here is the shape of it.
EU AI Act enters into force
The world's first comprehensive AI law starts its phased rollout.
First bans and AI literacy duties apply
Prohibited uses are outlawed and staff AI literacy becomes an expectation.
Data (Use and Access) Act receives Royal Assent
The UK's main data law reform is passed, with provisions phased in from 2026.
Rules for general-purpose AI apply
Obligations for large general-purpose models (the kind behind most chat tools) take effect.
New automated-decision rules and charitable soft opt-in
New UK GDPR Articles 22A to 22D on automated decisions, plus an easier soft opt-in for charity supporter emails.
AI transparency duties
Rules on labelling AI-generated content and being clear when people deal with AI begin to land.
High-risk AI obligations apply
The strictest duties, for higher-risk AI uses, come into force after a deferral.
Dates are indicative and simplified, and some have already moved. Current as of July 2026. We confirm the detail for your specific situation.
The laws that touch you
Four things worth knowing
You do not need to become a lawyer. You do need to know which rules apply and have a plan. These are the four that matter most for a UK charity right now.
UK GDPR and the ICO
The seven data protection principles apply in full to AI. Special category data such as health and religious belief, and any safeguarding information, carry a higher bar. The ICO guidance on AI and data protection is the reference point.
Data (Use and Access) Act 2025
Now in force. It changes the rules on solely automated decisions (new UK GDPR Articles 22A to 22D, from February 2026) and adds a charitable purposes soft opt-in for supporter emails. Charities must also have a data protection complaints process in place.
EU AI Act
A risk-based law that can reach UK charities whose AI outputs are used in the EU. Rules for general-purpose AI already apply, transparency duties land through 2026, and high-risk obligations follow in 2027. Knowing your risk tier is the starting point.
ISO/IEC 42001 and the Charity Governance Code
ISO 42001 is the international standard for managing AI responsibly and is increasingly used as a procurement signal. The 2025 Charity Governance Code refresh now expects charities to have a technology and AI policy.
The simplest rule that keeps you safe
What you can and cannot put into AI
Most AI risk in a charity comes down to one question: what data went into the tool? This traffic light is the single most useful thing your team can keep on the wall. We tailor it to your organisation as part of the work.
GREEN
Go ahead
No personal data. Safe to use.
AMBER
Stop and strip it first
Remove names and details, then proceed.
RED
Never, in a consumer account
This must not go into a free or personal AI tool.
The gut check: would you be relaxed if this exact text appeared in an email to someone outside the organisation? If not, it does not belong in a consumer AI account.
A quick self-check
Are you ready?
If you paused on any of these, that is where we start. None of it needs to be daunting, and most of it can be in place in weeks, not months.
Ethical procurement and use
Choosing AI tools responsibly
Not every AI tool is a safe home for your data. Before you buy or roll something out, these are the questions to ask. We build this into a checklist your team can reuse for every new tool.
How I help
What we put in place
Six building blocks of a responsible AI approach. We start with where you are, and do the ones that matter most first.
AI strategy and roadmap
A clear, realistic plan for where AI helps your mission and where it does not. Priorities, sequencing, and the questions your board should be asking.
AI policy and acceptable use
A written policy your team actually understands: what AI can be used for, what it must never touch, and who signs off. Aligned with the Charity Governance Code.
Data protection and safeguarding
Where the line sits for personal, special category and safeguarding data under UK GDPR, and how to keep the people you serve safe.
Governance and oversight
A named owner, a simple board reporting rhythm, and trustee oversight that meets your duties under the Charities Act 2011.
Ethical procurement
How to choose and contract AI tools responsibly: data residency, training on your data, processing agreements, and vendor due diligence.
Team fluency and training
Hands-on training so staff use AI confidently and safely, built around your real work, not a policy that sits in a drawer.
Yours to keep
What you walk away with
Working documents your team will actually use, not a report that sits on a shelf.
A tailored AI policy
A clear, plain-English policy for your organisation, ready to adopt and share with staff and trustees.
A responsible-use guide
The green, amber and red guide above, adapted to your organisation, so staff make safe choices without asking every time.
A governance model
A named owner, a light board reporting rhythm, and a review schedule that meets your trustee duties.
A procurement checklist
The questions to ask any AI vendor before you buy, so due diligence is consistent and documented.
Staff training and toolkit
Hands-on training and a take-home toolkit of prompts and workflows built around your real work.
Work with me
Talk to me about AI strategy and governance
I am a strategist with 25 years across philanthropy, policy and investment, and the founder of Arakura. I work with charities and nonprofits to put AI strategy and governance in place, and to train teams to use it well.
Looking for hands-on team training instead? See AI training for your team ↑
A short call to understand where you are and what you need. No cost, no obligation.
Latest insights & updates
On the subjects that matter here
Articles, posts and thinking on AI training, strategy, policy and governance — only the pieces relevant to organisations, not everything I publish.
We use cookies to improve your experience and analyse site traffic. Non-essential cookies (e.g. analytics) are only set with your consent. Cookie Policy