Across research, evaluation, facilitation, policy and organisational settings, AI is already being used for early thinking, drafting, analysis, synthesis and communication. These uses can be valuable. They can extend what individuals and small teams are able to do, surface alternatives and make difficult material easier to work with.
But a useful output does not necessarily mean that the wider work has improved. AI can also shape how a problem is framed, which perspectives remain visible, where interpretation occurs, how participation is represented and how much scrutiny a polished result receives.
The practical question is therefore not simply whether to use AI, or whether an output appears satisfactory. It is whether AI is contributing to better work, for whom, and under what conditions. That requires attention to the inquiry, participation, judgement, responsibility and learning that surround its use.
The pages and essays below offer different starting points. They include practical resources for research and reflective practice, wider ethical and organisational perspectives, guidance for developing AI policies, and essays on interpretation, participation, evaluation and organisational learning.
I also describe how I use AI in developing this site’s essays and resource pages on the site structure and navigation page.
Choose a starting point
Use this section to find a useful starting point, whether you are new to AI, thinking about wider ethical questions, using AI as a thought partner, developing prompts, or exploring how AI changes shared work.
- New to this collection
Start with What would better work with AI look like?. It sets out the central question running through the collection: how AI might support professional work without separating useful outputs from inquiry, participation, interpretation and responsibility. - Looking for practical guidance and resources
Go to Using AI in research and practice for practical guidance and selected resources on preparation, reflection, analysis, qualitative research, writing, evaluation and responsible use. - Wanting to understand the wider context
Start with Seeing the wider ethical picture around AI development and use for a framing of AI ethics across structural, organisational and practice levels. Then explore AI in context: the wider picture for a curated set of reports and resources on AI’s wider social, environmental and organisational implications. - Using AI for reflective practice or writing
Read AI as a thought partner and AI prompts for shared thinking for practical reflections on using AI to support early thinking, drafting and preparation. - When AI is entering shared work
Explore Working with AI in the room, the Authority of the first synthesis, and AI can produce the record of participation without the participation for reflections on authorship, framing, participation and collective judgement. AI in place-based practice considers how these shifts play out in longer-term collaborative settings. - When your team or organisation is deciding where to begin
Visit Working with AI: where to begin for a practical starting point for conversations about purpose, use, review and responsibility. - Evaluating whether AI is improving the work
Read How do we know whether AI is improving the work? for an approach that connects output checks with purpose, evidence, shared judgement and adaptation. - Developing AI policy or guidance
Start with Developing an organisational AI policy for a concise set of questions about purpose, current use, responsibility, scrutiny and review. Then explore AI policy templates and guidance for adaptable examples, staff guidance, governance frameworks and assessment tools.
Browse all essays and resource pages
The starting points above help visitors find the part of this section that best matches their current interests. The essays and resource pages below provide a fuller map of the material.
Essays on AI and professional practice
These essays explore what happens when generative AI becomes part of professional and collaborative work. The opening essay introduces the wider question of what better work with AI would mean. The essays that follow examine how that question plays out in individual practice, shared work, interpretation, participation, evaluation, organisational learning, and finally within the wider ethical landscape. They can be read individually or followed as a developing series.
What would better work with AI look like?
Introduces the central question running through the collection: whether AI is contributing to better work, rather than simply producing useful-looking outputs. It considers purpose, participation, interpretation, responsibility and the organisational conditions that help people question practice and learn where AI should continue, change or stop.
AI as a thought partner: reflections on collaborative practice and systems work
Begins with individual practice. Drawing on personal experience, this essay considers how AI can support early thinking, test alternative framings and help prepare ideas and material for further work, while leaving interpretation, relationships and responsibility with people.
AI prompts for shared thinking: a light framework for purposeful prompting
Builds on the previous essay by introducing a simple prompting framework that keeps purpose, context, relationships and intended use in view. It treats prompts as part of a wider thinking process rather than instructions for producing a finished answer.
Working with AI in the room: authorship, responsibility, and collective judgement
Moves from individual use into shared work. It examines what changes when AI-generated material enters meetings, workshops, strategy discussions and collaborative writing, including questions about authorship, trust, responsibility and how collective sense-making is held.
AI and the authority of the first synthesis
Looks more closely at the power of early interpretation. It considers how an AI-assisted summary, set of themes or draft strategy can shape what a group notices and treats as settled, and why the first synthesis needs to be examined rather than simply polished.
AI can produce the record of participation without the participation
Extends the argument into participation and representation. It asks what is missing when AI can generate transcripts, themes and reports that resemble the products of engagement, but not the influence, relationships, disagreement and learning through which meaningful participation occurs.
How do we know whether AI is improving the work?
Turns to evaluation. It shows why output checks alone are insufficient. Drawing on established approaches to indicators, evaluative rubrics and monitoring, evaluation and learning, it explores how AI-supported work can be followed into later discussion, decisions and consequences.
Working with AI: where to begin
Offers a practical starting point for organisations, teams and programmes wanting to engage with AI more deliberately. It sets out five areas to work through, beginning with current use and moving through purpose, participation, responsibility and regular review rather than assuming that a complete framework must come first.
AI in place-based practice: what is shifting
Applies the collection’s ideas to place-based collaboration, where relationships, long timeframes and multiple organisations make questions of judgement, participation and learning especially important. It considers how AI may influence preparation, issue framing and collective sense-making in these settings.
Seeing the wider ethical picture around AI development and use
Concludes the collection by placing the practical questions explored in the earlier essays within a broader ethical landscape. It distinguishes structural, organisational and practice-level concerns, showing why different perspectives on AI often arise because they address different parts of the same system, and why each level calls for different kinds of responsibility and response.
Resource pages
These resource pages bring together curated and annotated links to practical guidance, reports, frameworks and further reading, alongside reflections from my own practice.
Using AI in research and practice
Brings together practical guidance and curated resources on using AI in research, evaluation and other forms of professional practice. The page focuses on inquiry, interpretation, writing, qualitative research and responsible use, while keeping human judgement and responsibility at the centre.
AI in context: the wider picture
Explores the organisational, social and environmental conditions surrounding AI use. It brings together resources on adoption, governance, infrastructure and environmental impacts, complementing the practice-focused essays elsewhere in this section.
Developing an organisational AI policy
Offers a concise set of questions to help smaller organisations develop practical AI policies or internal guidance. It covers why guidance is needed, what is already happening, what should be included, which uses deserve greater scrutiny, whose material is involved, who should help shape the guidance, and how it will be reviewed as practice changes.
AI policy templates and guidance
Brings together selected policy templates, staff guidance, governance frameworks and assessment tools that organisations can compare and adapt. It is intended as a practical companion to the organisational AI policy page rather than a single model to follow.
Quick answers to common questions
How can I use AI without losing my own judgement?
AI can support early thinking, drafting and analysis, but judgement may narrow when an AI-generated summary or synthesis becomes the starting point for everyone else. It helps to compare outputs with source material, test alternative framings and keep interpretation open long enough to question what has been carried forward and what has been left out.
Do I need to tell clients, participants or readers that I used AI?
Disclosure matters most when AI has materially shaped analysis, interpretation, advice, published writing or how other people’s views are represented. Minor support with spelling or formatting may not require the same explanation. A useful disclosure says what AI contributed, not simply that it was used. It should also remain clear who has reviewed the work, accepts responsibility for it and is prepared to stand behind the final judgement.
Is it safe to put interview transcripts or participant submissions into an AI tool?
Not automatically. A paid service may offer stronger privacy commitments than a free public tool, but that does not make it suitable for confidential or sensitive material. Check what consent covered, whether the provider stores or reuses data, and what organisational, contractual, ethical or data-governance requirements apply. Existing ethics approvals, funder conditions or data-sharing agreements may already limit what is permitted. De-identification can reduce some risks, but context, relationships or community-held knowledge may still remain identifiable.
Is it appropriate to use AI to summarise consultation submissions or workshop discussions?
It can be useful, but the summary should not be treated as a neutral record. AI may flatten disagreement, omit minority perspectives or make tentative views appear settled. Where the summary represents participants, they may need an opportunity to see how their views have been represented, question the interpretation and correct important omissions before it shapes later discussion or decisions. The original material should remain available for checking, and responsibility for the final interpretation stays with the people leading the process.
How can we tell whether AI is actually improving our work?
A satisfactory output is only one source of evidence. Organisations should also examine what changed around the work: whether inquiry became stronger or narrower, whether important perspectives remained visible, whether checking increased, whether decisions improved and who benefited or carried additional risk. Useful review questions include what became better, for whom, and what should continue, change or stop.
Does a small organisation need an AI policy?
Not every small organisation needs a lengthy standalone policy. It may need a short set of shared expectations covering acceptable use, privacy, checking, disclosure, responsibility and review. The value often lies less in the document than in making current practice visible and agreeing where greater care or outside advice is needed. The Developing an organisational AI policy page provides a practical starting point.
These resources are for practitioners, policymakers, facilitators and others looking to use AI in ways that are inclusive, reflective and grounded in real practice. The aim is not greater AI use in itself, but better work. These resources support practitioners and organisations to consider where AI is useful, what needs greater scrutiny, and how inquiry, participation, judgement and learning can remain visible as practice changes.
[* Image by Jintana / Adobe Stock]