How I use AI at Thorp Coaching
Artificial intelligence is increasingly built into many of the tools we use in everyday life and work. It can be useful, but I also recognise that people may have understandable questions or concerns about how it is used—particularly in the context of confidential coaching conversations.
I believe you should be able to understand where I use AI, why I use it and where I deliberately do not.
AI supports some of the practical and creative work around my coaching practice. It does not replace my professional judgement, accountability or the human relationship at the centre of the coaching itself.
How I use AI-enabled tools
I use a range of AI-enabled tools, including ChatGPT, Google Gemini, Canva and video-editing software.
I may use these tools to:
Copy-edit material I have written.
Structure non-confidential elements of proposals and programme outlines.
Develop and refine generic workshop materials.
Support research using publicly available information.
Help me organise my thinking and communicate it more clearly.
Create and refine marketing, design and video content.
Support routine administrative tasks.
Support business development and professional relationship management.
I use these tools selectively rather than automatically. I review and edit anything produced with their assistance and remain responsible for the final work. I do not treat AI-generated material as automatically accurate, original or appropriate.
For routine business administration, business development and professional relationship management, I may use AI-enabled tools — including through defined connections to business email or documents — to process limited professional information such as a person’s name, organisation, role, relevant business correspondence and relationship status. I limit access to the information necessary for the specific task and review all outputs before taking action. This does not include confidential coaching content, coaching notes, assessments, sensitive personal information or information identifying someone as a coaching client.
I use a paid version of ChatGPT and have disabled the setting that allows my new conversations to be used to improve OpenAI’s models. Where I use AI with limited professional information, I also use the available privacy and access controls and delete working material when it is no longer required.
AI-assisted structuring of generalised coaching themes
On occasion, I may use AI to help structure a brief summary after a coaching session. Before using an AI tool, I reduce my notes to broad themes and agreed actions and remove names, employers, exact roles, colleague names, quotations, health or diagnostic information and other details that could reasonably identify the client, another person or an organisation.
I do not upload raw coaching notes, recordings, transcripts or detailed accounts of coaching conversations to general-purpose AI tools. AI is used only to organise or improve the wording of themes and actions that I have already identified. It does not interpret the session, create coaching observations, assess the client or decide how I should coach them.
Any final note or summary is reviewed, amended and approved by me before it is stored or shared.
Information I do not provide to general-purpose AI tools
In my work on behalf of Thorp Coaching, I do not provide general-purpose AI tools with:
Raw or identifiable confidential coaching conversations.
Client notes, recordings or transcripts, except where a separate transcription service is used with the client’s active agreement as described below.
A client’s name, employer or exact role in connection with coaching information.
Sensitive personal, health, diagnostic, accessibility or assessment information about a client or coachee.
Confidential organisational, board, legal, financial, employment or commercial information.
Exact quotations or detailed, distinctive personal or professional circumstances that could reasonably identify a client, another person or an organisation.
Psychometric assessments, sponsor feedback or performance information.
Where I use examples to help me think or write, I fictionalise, remove or generalise identifying and sensitive details.
How I do not use AI
I do not use AI to:
Deliver coaching to clients.
Diagnose, psychologically assess or profile a client.
Make judgements about someone’s performance, potential, suitability for a role or the coaching support they should receive.
Decide what information should be shared with an organisational sponsor.
Generate assessments, sponsor reports or organisational conclusions from raw or identifiable confidential client material.
Replace my professional judgement or the relationship between coach and client.
AI may support limited practical work around coaching. It is not used to tell me how to coach you.
AI transcription and note-taking
I currently use Otter as an optional AI transcription and note-taking tool for some meetings and coaching sessions.
Some clients find having a transcript or summary helpful. It can allow them to remain more present in the conversation rather than trying to listen, think and take detailed notes at the same time. This can be particularly useful for people who prefer to process information after a session.
Other clients may feel uncomfortable with an AI tool recording or processing a coaching conversation. That is equally understandable.
Otter is only used where you have actively opted in. Its use is entirely optional, and choosing not to use it will not affect the coaching you receive.
Even if you have previously opted in, you can ask for Otter to be removed before or at any point during a session. We will simply continue without AI transcription, and I will take my usual brief coaching notes.
Why I am moving away from Otter
Otter is a US-based provider and data may be processed outside the UK.
Its published privacy information states that it may use de-identified audio recordings and transcripts to train and improve its technology. However, the setting that allows conversations from my account to be shared for training and product improvement has always been disabled.
Even with that setting switched off, I have decided that I would prefer to use a provider offering UK or EU data storage and privacy arrangements that are better aligned with the confidential nature of coaching.
This decision is not in response to a data breach or any known problem with client information. It is a proactive, belt-and-braces decision based on the standard I want to maintain as the technology and our understanding of it continue to develop.
I have cancelled my Otter subscription, which will remain active until 21 March 2027.
Until then, clients are very welcome to continue using Otter in their sessions if they find the transcripts or notes useful. Its use will remain entirely optional.
Once I have selected a replacement provider, I will update this statement to explain:
Which provider I use.
Where recordings and transcripts are stored.
Who can access them.
How long they are retained.
Whether any information is used for model training.
Existing Otter transcripts
If you would like to keep an existing transcript, please download it or ask me for a copy by 28 February 2027.
Providing you with a copy will not mean that I retain my own copy.
Unless we have agreed a specific reason to retain it, I will delete all recordings and transcripts from my Otter account by 20 March 2027.
After that date, I will no longer be able to retrieve them.
Transcripts may contain confidential or personal information. If you choose to keep a copy, please consider where and how you store it.
Human accountability
AI may help with some of the activities surrounding my work, but it does not replace the deep listening, challenge, professional experience, intuition, ethical judgement and human relationship involved in coaching.
Any coaching observations, questions, recommendations, written summaries or organisational reports are considered and approved by me. I remain professionally accountable for the service I provide.
Where AI has assisted with the organisation or wording of generalised themes and actions, I check the result against my own notes and judgement. AI-generated wording is not treated as a factual observation or professional conclusion merely because a tool has produced it.
Using AI proportionately
I am conscious that AI is not environmentally neutral. It depends on data centres, computing equipment and energy, and I believe responsible use should consider environmental impact alongside privacy, confidentiality and accuracy.
I therefore do not use AI automatically simply because it is available.
I use it selectively for defined tasks where I believe it adds clear value. I avoid unnecessary or repeated generation where there is no clear purpose, and I continue to use my own judgement or simpler tools where they can do the job equally well.
This is not a claim that my use of AI has no environmental impact. It is a commitment to use it thoughtfully and proportionately and to keep reviewing my approach as the technology develops.
Security, retention and review
I review the privacy, security and contractual arrangements of AI-enabled tools before approving them for business use. I use available settings to prevent my content being used for model training, restrict access to my accounts and apply appropriate authentication controls.
Working prompts and outputs containing limited professional information are retained only for as long as needed for the task. Where AI has assisted with generalised coaching themes and actions, the working conversation is deleted from my active AI account when the final note has been completed and no later than 30 days afterwards. The final, coach-approved record is stored within my secure client-record system and follows my normal retention policy. Provider backups or legally required retention may continue for a limited period after deletion from my account.
I keep my use of AI and the providers I use under review as technology, professional guidance and data-protection requirements develop.
Alignment with professional standards
My approach is informed by the International Coaching Federation’s Code of Ethics and AI Coaching Framework, as well as the EMCC Global Code of Ethics and Digital Ethics Guidelines.
Together, these professional standards emphasise:
Transparency about how technology is used.
Informed agreement where technology records or processes coaching conversations.
Protection of client confidentiality, privacy and security.
Responsible storage and disposal of records.
Appropriate quality and professional conduct in technology-enabled coaching.
Continuing human judgement and accountability.
Respect for copyright and intellectual property.
These standards guide how I assess the tools I use and how I review my practice as technology develops.
Questions and choices
You are welcome to ask how AI is being used in relation to your work at any time.
Where a tool would record, transcribe or otherwise process the content of a coaching conversation, you will always be asked to opt in actively. You can also change your mind before or during a session.
If I ever propose using a tool to process identifiable or sensitive coaching material beyond the optional transcription arrangement described above, I will first explain the provider, purpose, information involved, storage and retention arrangements and ask you to opt in actively. Declining will not affect the coaching you receive, and you may change your mind before or during the relevant processing.
This statement will be reviewed and updated as my tools, professional guidance and practice develop.
Last updated: September 2026.