Professional Ethics — a three-week foundation

Synthesis, citation, and the borrowing that felt small enough not to mention.

• ✓
  learnRequired

  Attribution is four questions, not one

  Read the brief~2 min

  Almost every argument about attribution is really an argument about one of four different questions, and people lose it by answering the wrong one. Is this mine? Is it mine with a citation? Is it mine with the machine's help declared? Or is it not mine to ship at all?

  The four answers are not a severity ladder

  It is tempting to read them as increasing caution, with “just cite it” as the safe default. That is wrong in a way that costs you. A conclusion you reached by holding six sources side by side, which none of them states, is YOURS — and hanging a citation on it hands the responsibility to someone who never made the claim. Over-citation reads as rigour and is often a refusal to own an argument.

  The fourth answer is the only one that stops the work, and it is the one people reach for last. Some material cannot be shipped under any label: a colleague's slide built from another client's situation, a figure you cannot source, a case study a model invented about a real company. No credit line rescues those, because attribution was never the problem.

  The test that actually works

  • Would a competent peer need a reference to trust this? If not, citing it is padding.
  • If the source turned out to be wrong, whose error is it? That answer tells you whether a citation is doing real work.
  • Could I defend this in a room with no tool and no document in reach? If not, the understanding is not yours yet.
  • Is the problem WHO wrote it, or WHOSE SITUATION it describes? The second is confidentiality, and it does not respond to attribution at all.

  The reason this matters at the start of the programme is that everything later depends on it. You cannot decide what to declare until you can say whose work you are holding, and you cannot stand behind a commitment made on evidence you have not examined.

  Go deeper — with your own AICopy-paste prompt

  The brief above is the spine. This is a full study prompt for whichever assistant you already use — it carries your programme, this module and what the brief just covered, so the lesson comes back tailored rather than generic. Pick how you want it, edit anything, then copy.

  WrittenInfographicBoth

  You are an experienced control-risk practitioner and an exceptional teacher. Teach me one topic properly — the way a senior colleague would over a long coffee, not the way a textbook would. MY CONTEXT - Programme: Professional Ethics — a three-week foundation (3 weeks, ~4h a week, self-paced). - Level: Anyone who hands work to someone who will rely on it. No domain is assumed and none is needed: a graduate on a first deliverable, a founder signing a first contract, and a practitioner twenty years in who has never been asked any of this out loud are all in scope, and will find different parts of it uncomfortable. - Week 1 of 3: Whose work is this? — Attribution is not one question but four, and only one of the answers stops the work. Establish whose work you are holding before you decide how to label it — and what changes when a machine helped. - Module 1.1: Owning a claim, and borrowing one - Today's topic: Attribution is four questions, not one - I am studying alone, applying this to my own organisation, and nothing I produce is graded. Assume I am a working professional, not a student: skip the motivational framing and the history lesson unless the history changes what I should do. WHAT I ALREADY HAVE (do not simply repeat this — go past it) - The four answers are not a severity ladder - The test that actually works - Would a competent peer need a reference to trust this? If not, citing it is padding. - If the source turned out to be wrong, whose error is it? That answer tells you whether a citation is doing real work. - Could I defend this in a room with no tool and no document in reach? If not, the understanding is not yours yet. - Is the problem WHO wrote it, or WHOSE SITUATION it describes? The second is confidentiality, and it does not respond to attribution at all. WHAT I WANT FROM YOU 1. The core idea in five sentences, in plain language, including the one distinction practitioners most often get wrong here. 2. Why it matters in practice: what actually goes wrong when this is done badly, with two realistic short scenarios — one financial services, one outside it. 3. The mechanics, step by step, at the level of detail I could follow at my desk tomorrow. Show the judgement calls, not just the procedure, and say what a good answer looks like versus a defensible-but-weak one. 4. At least one fully worked example with realistic specifics — names, numbers, evidence, dates. Show your reasoning as you go, including where you would push back on your own conclusion. 5. The five most common mistakes, each with the tell that gives it away and the correction. 6. How this connects to the rest of control risk: what it depends on upstream and what breaks downstream if I get it wrong. 7. What competent practitioners genuinely disagree about here, and how I should decide for my own context. 8. A short glossary of any term you use that a regulator, an auditor and a business owner would each define differently. 9. Six self-test questions that require judgement rather than recall — hold the answers back until I have attempted them, then mark my reasoning, not just my conclusion. OUTPUT FORMAT: written explanation. Long-form prose with headed sections, worked examples and tables where a table genuinely helps. No images. HOW TO WORK WITH ME - Ask me up to three questions about my sector, my organisation's size and my current maturity BEFORE you start, then tailor everything to my answers. - Prefer concrete specifics over hedged generalities. If something depends on context, say which context and why. - Flag anything that is genuinely contested, jurisdiction-specific or changing, and tell me what would make you change your answer. - If I have got something wrong in what I said I already know, correct me directly. - Do not invent standards, clause numbers, statistics or citations. If you are unsure of a reference, say so plainly instead. - After the lesson, quiz me the way this module's drill would (whose-words): give me a situation, make me decide, then challenge the decision I make. - Finish by helping me plan my deliverable for this module: A provenance note on one real deliverable. Ask what my process is, then outline what mine should contain and what would make it weak.

  Copy promptYours to edit — changes stay in this box

• ✓
  playRequiredTicks itself

  Whose Words Are These?

  Twelve things you are about to put your name on, including three where the honest answer stops the work. It scores your reflex rate in both directions — including the cautious one, because citing to avoid owning a claim is also a miss.

  Open the drill

  Ticks itself when the drill records a result

• ✓
  applyRequired

  A provenance note on one real deliverable

  Open the brief

  Take one real deliverable you have sent in the last month — a report, a deck, a pull request, a proposal. Not your best one; the most ordinary one you can find, because the ordinary ones are where habits live.

  Write a provenance note for it. Nobody else needs to see this, and it should take twenty minutes rather than an afternoon. The point is to find out whether you can do it at all.

  What to produce

  • What the deliverable was, who relied on it, and what they did next.
  • Every element that did not originate with you: text, structure, figures, code, images, framing. List them even where the answer is obviously fine.
  • For each: which of the four answers applies, and what the deliverable actually said about it at the time.
  • The gap. Anything where the label you gave it and the label it needed are different — and whether anyone would have known.
  • One change you will make to the next one. One, specific, not a resolution.

  Template · provenance-note
• ✓
  checkOptional

  Where you cited to avoid owning a claim

  Open the prompts

  1. Find the place in that deliverable where you added a citation you did not need. What were you protecting yourself from?
  2. Was there a conclusion you softened into someone else's words because owning it felt exposed? What would it have said if you had owned it?
  3. If a colleague had produced the same document, is there anything in it you would have questioned?

The assistance barely changes. The promise you already made to the reader changes everything.

• ✓
  learnRequired

  Read the undertaking, not the tool

  Read the brief~2 min

  There is no single right answer to “should I declare that I used AI”, and looking for one is the mistake. The assistance barely changes from setting to setting. What changes is the promise you had already made to whoever reads the output.

  Four settings, four different obligations

  • An internal note from your own meeting: nothing about the reader's reliance changes. Declaring it trains people to skim declarations.
  • Work going out under someone else's name as well as yours: they are accepting responsibility, and cannot do that on information they do not have. Tell them; whether the client is told is then their call.
  • A client deliverable or a published byline: the reader has no other way to judge, so the declaration belongs in the artefact.
  • An exam condition, a confidentiality undertaking, a judgement about a person: these are not disclosure questions. Declaring does not fix them, because the problem is that you agreed to something else.

  Declaring everything is a way of declaring nothing

  A blanket policy of disclosing all assistance feels safest and is not. If a grammar fix carries the same notice as a generated analysis, the notice stops carrying information and readers stop reading it — which costs you exactly the cases where it mattered. The useful line is whether the machine changed what the work SAYS.

  The uncomfortable one is relying on a summary of a document you have not read. You are citing a source through an intermediary that cannot be cross-examined. That is worth declaring, and worth treating as a prompt to go and read the part your recommendation actually rests on.

  Go deeper — with your own AICopy-paste prompt

  The brief above is the spine. This is a full study prompt for whichever assistant you already use — it carries your programme, this module and what the brief just covered, so the lesson comes back tailored rather than generic. Pick how you want it, edit anything, then copy.

  WrittenInfographicBoth

  You are an experienced control-risk practitioner and an exceptional teacher. Teach me one topic properly — the way a senior colleague would over a long coffee, not the way a textbook would. MY CONTEXT - Programme: Professional Ethics — a three-week foundation (3 weeks, ~4h a week, self-paced). - Level: Anyone who hands work to someone who will rely on it. No domain is assumed and none is needed: a graduate on a first deliverable, a founder signing a first contract, and a practitioner twenty years in who has never been asked any of this out loud are all in scope, and will find different parts of it uncomfortable. - Week 1 of 3: Whose work is this? — Attribution is not one question but four, and only one of the answers stops the work. Establish whose work you are holding before you decide how to label it — and what changes when a machine helped. - Module 1.2: Declaring the machine - Today's topic: Read the undertaking, not the tool - I am studying alone, applying this to my own organisation, and nothing I produce is graded. Assume I am a working professional, not a student: skip the motivational framing and the history lesson unless the history changes what I should do. WHAT I ALREADY HAVE (do not simply repeat this — go past it) - Four settings, four different obligations - An internal note from your own meeting: nothing about the reader's reliance changes. Declaring it trains people to skim declarations. - Work going out under someone else's name as well as yours: they are accepting responsibility, and cannot do that on information they do not have. Tell them; whether the client is told is then their call. - A client deliverable or a published byline: the reader has no other way to judge, so the declaration belongs in the artefact. - An exam condition, a confidentiality undertaking, a judgement about a person: these are not disclosure questions. Declaring does not fix them, because the problem is that you agreed to something else. - Declaring everything is a way of declaring nothing WHAT I WANT FROM YOU 1. The core idea in five sentences, in plain language, including the one distinction practitioners most often get wrong here. 2. Why it matters in practice: what actually goes wrong when this is done badly, with two realistic short scenarios — one financial services, one outside it. 3. The mechanics, step by step, at the level of detail I could follow at my desk tomorrow. Show the judgement calls, not just the procedure, and say what a good answer looks like versus a defensible-but-weak one. 4. At least one fully worked example with realistic specifics — names, numbers, evidence, dates. Show your reasoning as you go, including where you would push back on your own conclusion. 5. The five most common mistakes, each with the tell that gives it away and the correction. 6. How this connects to the rest of control risk: what it depends on upstream and what breaks downstream if I get it wrong. 7. What competent practitioners genuinely disagree about here, and how I should decide for my own context. 8. A short glossary of any term you use that a regulator, an auditor and a business owner would each define differently. 9. Six self-test questions that require judgement rather than recall — hold the answers back until I have attempted them, then mark my reasoning, not just my conclusion. OUTPUT FORMAT: written explanation. Long-form prose with headed sections, worked examples and tables where a table genuinely helps. No images. HOW TO WORK WITH ME - Ask me up to three questions about my sector, my organisation's size and my current maturity BEFORE you start, then tailor everything to my answers. - Prefer concrete specifics over hedged generalities. If something depends on context, say which context and why. - Flag anything that is genuinely contested, jurisdiction-specific or changing, and tell me what would make you change your answer. - If I have got something wrong in what I said I already know, correct me directly. - Do not invent standards, clause numbers, statistics or citations. If you are unsure of a reference, say so plainly instead. - After the lesson, quiz me the way this module's drill would (declare-it): give me a situation, make me decide, then challenge the decision I make. - Finish by helping me plan my deliverable for this module: The AI-use line you would actually publish. Ask what my process is, then outline what mine should contain and what would make it weak.

  Copy promptYours to edit — changes stay in this box

• ✓
  playRequiredTicks itself

  Declare It

  One task, twelve settings, four answers — and three of the answers are not about disclosure at all. It counts your one-rule rate: applying a single policy across settings that promised different things.

  Open the drill

  Ticks itself when the drill records a result

• ✓
  applyRequired

  The AI-use line you would actually publish

  Open the brief

  Write the AI-use line you would actually be willing to publish, for the kind of work you do most. Not a policy for your organisation — one or two sentences you would put in a real deliverable this week.

  The test of it is whether it is specific enough to be falsifiable. “AI tools were used in the preparation of this document” passes no test at all, because it is true of everything and describes nothing.

  What to produce

  • The setting you are writing it for, and who relies on the output.
  • What you actually do: which tasks, at which stage, and where you stop.
  • The line itself, in the words you would publish.
  • What it does NOT cover, stated plainly — the honesty is in the exclusions.
  • Where you would put it, and who would see it.

  Template · ai-use-line
• ✓
  checkRequired

  The setting where your current practice would not survive being described out loud

  Open the prompts

  1. Name the setting where your current practice would not survive being described out loud to the person relying on it. Do not resolve it yet — just name it.
  2. Is there anything you have declared purely to protect yourself, which the reader gained nothing from?
  3. Have you ever relied on a summary of something you did not read, in work someone acted on? What would you do differently now, and what would that cost?

A number in a deliverable is a promise that someone could check it.

• ✓
  learnRequired

  Provenance is a property of the evidence, not of your confidence

  Read the brief~2 min

  A number in a deliverable is a promise that someone could check it. Most of the time nobody does, which is why the habit has to come from you rather than from being caught.

  Confidence is not provenance

  The dangerous claims are not the ones you are unsure about — you hedge those. They are the ones you are confident about and cannot source: the statistic you remember reading, the figure everyone in your field repeats, the rule of thumb you were taught. Confidence is a fact about you. Provenance is a fact about the evidence, and only one of them survives a challenge.

  • An undated screenshot is not evidence of a control operating; it is evidence that something once looked like that.
  • A figure you cannot find is not saved by “approximately”. You can keep the knowledge and drop the false precision: say what you know and how you know it.
  • A model's fluent output about a named party is the highest-risk kind, because it reads exactly like a checked fact.
  • Extraction you observed is different in kind from extraction you were handed. Both can be fine; only one is yours to vouch for.

  This module reuses a drill written for auditors, and deliberately. The reliability of evidence is not a specialist concern — it is the same question a founder faces in a pitch deck and a graduate faces in their first memo, asked by people whose job is to be strict about it.

  Go deeper — with your own AICopy-paste prompt

  The brief above is the spine. This is a full study prompt for whichever assistant you already use — it carries your programme, this module and what the brief just covered, so the lesson comes back tailored rather than generic. Pick how you want it, edit anything, then copy.

  WrittenInfographicBoth

  You are an experienced control-risk practitioner and an exceptional teacher. Teach me one topic properly — the way a senior colleague would over a long coffee, not the way a textbook would. MY CONTEXT - Programme: Professional Ethics — a three-week foundation (3 weeks, ~4h a week, self-paced). - Level: Anyone who hands work to someone who will rely on it. No domain is assumed and none is needed: a graduate on a first deliverable, a founder signing a first contract, and a practitioner twenty years in who has never been asked any of this out loud are all in scope, and will find different parts of it uncomfortable. - Week 1 of 3: Whose work is this? — Attribution is not one question but four, and only one of the answers stops the work. Establish whose work you are holding before you decide how to label it — and what changes when a machine helped. - Module 1.3: Evidence you can stand behind - Today's topic: Provenance is a property of the evidence, not of your confidence - I am studying alone, applying this to my own organisation, and nothing I produce is graded. Assume I am a working professional, not a student: skip the motivational framing and the history lesson unless the history changes what I should do. WHAT I ALREADY HAVE (do not simply repeat this — go past it) - Confidence is not provenance - An undated screenshot is not evidence of a control operating; it is evidence that something once looked like that. - A figure you cannot find is not saved by “approximately”. You can keep the knowledge and drop the false precision: say what you know and how you know it. - A model's fluent output about a named party is the highest-risk kind, because it reads exactly like a checked fact. - Extraction you observed is different in kind from extraction you were handed. Both can be fine; only one is yours to vouch for. WHAT I WANT FROM YOU 1. The core idea in five sentences, in plain language, including the one distinction practitioners most often get wrong here. 2. Why it matters in practice: what actually goes wrong when this is done badly, with two realistic short scenarios — one financial services, one outside it. 3. The mechanics, step by step, at the level of detail I could follow at my desk tomorrow. Show the judgement calls, not just the procedure, and say what a good answer looks like versus a defensible-but-weak one. 4. At least one fully worked example with realistic specifics — names, numbers, evidence, dates. Show your reasoning as you go, including where you would push back on your own conclusion. 5. The five most common mistakes, each with the tell that gives it away and the correction. 6. How this connects to the rest of control risk: what it depends on upstream and what breaks downstream if I get it wrong. 7. What competent practitioners genuinely disagree about here, and how I should decide for my own context. 8. A short glossary of any term you use that a regulator, an auditor and a business owner would each define differently. 9. Six self-test questions that require judgement rather than recall — hold the answers back until I have attempted them, then mark my reasoning, not just my conclusion. OUTPUT FORMAT: written explanation. Long-form prose with headed sections, worked examples and tables where a table genuinely helps. No images. HOW TO WORK WITH ME - Ask me up to three questions about my sector, my organisation's size and my current maturity BEFORE you start, then tailor everything to my answers. - Prefer concrete specifics over hedged generalities. If something depends on context, say which context and why. - Flag anything that is genuinely contested, jurisdiction-specific or changing, and tell me what would make you change your answer. - If I have got something wrong in what I said I already know, correct me directly. - Do not invent standards, clause numbers, statistics or citations. If you are unsure of a reference, say so plainly instead. - After the lesson, quiz me the way this module's drill would (rely-or-reject): give me a situation, make me decide, then challenge the decision I make. - Finish by helping me plan my deliverable for this module: Trace three claims in your last deliverable to source. Ask what my process is, then outline what mine should contain and what would make it weak.

  Copy promptYours to edit — changes stay in this box

• ✓
  playRequiredTicks itself

  Rely or Reject — an evidence-reliability drill

  Eighteen pieces of evidence with the provenance you would actually have, from an undated screenshot to an extraction you watched being run. Written for auditors, and the judgement is general: what can you stand behind, and what are you merely confident about.

  Open the drill

  Ticks itself when the drill records a result

• ✓
  applyRequired

  Trace three claims in your last deliverable to source

  Open the brief

  Take your last substantive deliverable and pick three claims from it. One number, one statement about how something works, and one claim about what someone else did or said.

  Trace each to source. Not to where you got it — to source. Give yourself a time limit of ten minutes per claim, and record what happens when you run out.

  What to produce

  • The three claims, quoted as they appeared.
  • For each: the actual source, or the point at which the trail stopped.
  • How long each took, and whether the ten minutes was enough.
  • Which of the three you would now rewrite, and how.
  • One habit that would have made this easier — something you do while drafting, not a review step you will skip.

  Template · claim-trace

Absorb, record, raise, decline — four different answers, and refusing everything is wrong.

• ✓
  learnRequired

  Scope is a record, and silence erases it

  Read the brief~2 min

  Scope is not a document. It is a record, and the thing that destroys a record is not refusal or argument — it is generosity. Every small unrecorded favour is a true thing that becomes invisible, and the invisible things are what make the budget conversation in week six impossible to have honestly.

  Four answers, and refusing everything is wrong

  • ABSORB. Minutes, no precedent, and the job is better for it. Logging these would tell the client that everything now has a meter running, and you would spend goodwill you need for the requests that cost something.
  • RECORD. Worth doing and too big to vanish. Not about billing — about the record being true when someone asks where the time went.
  • RAISE. It changes what the engagement is. Starting it and mentioning it later hands the decision to everyone else after it has already been taken.
  • DECLINE. Not a scope conversation at all, and agreeing was never within your gift.

  The generous-looking failure

  The error most people make is not obstruction, it is silent absorption: doing the extra work, well, without a word. It reads as goodwill and it removes the evidence. When the budget runs out nobody can see where it went, so the next proposal repeats the same estimate and the same person absorbs the same gap.

  The opposite failure is real too, and this week's drill scores it: declining is wrong three times out of twelve. A professional who treats every question as a change request is running a worse engagement and protecting nothing.

  Go deeper — with your own AICopy-paste prompt

  The brief above is the spine. This is a full study prompt for whichever assistant you already use — it carries your programme, this module and what the brief just covered, so the lesson comes back tailored rather than generic. Pick how you want it, edit anything, then copy.

  WrittenInfographicBoth

  You are an experienced control-risk practitioner and an exceptional teacher. Teach me one topic properly — the way a senior colleague would over a long coffee, not the way a textbook would. MY CONTEXT - Programme: Professional Ethics — a three-week foundation (3 weeks, ~4h a week, self-paced). - Level: Anyone who hands work to someone who will rely on it. No domain is assumed and none is needed: a graduate on a first deliverable, a founder signing a first contract, and a practitioner twenty years in who has never been asked any of this out loud are all in scope, and will find different parts of it uncomfortable. - Week 2 of 3: What did you promise? — Commitments are made in passing and paid for in full. Find what you actually agreed to, notice the requests that quietly changed it, and learn what a slipping date costs while you decide whether to mention it. - Module 2.1: What you actually agreed to - Today's topic: Scope is a record, and silence erases it - I am studying alone, applying this to my own organisation, and nothing I produce is graded. Assume I am a working professional, not a student: skip the motivational framing and the history lesson unless the history changes what I should do. WHAT I ALREADY HAVE (do not simply repeat this — go past it) - Four answers, and refusing everything is wrong - ABSORB. Minutes, no precedent, and the job is better for it. Logging these would tell the client that everything now has a meter running, and you would spend goodwill you need for the requests that cost something. - RECORD. Worth doing and too big to vanish. Not about billing — about the record being true when someone asks where the time went. - RAISE. It changes what the engagement is. Starting it and mentioning it later hands the decision to everyone else after it has already been taken. - DECLINE. Not a scope conversation at all, and agreeing was never within your gift. - The generous-looking failure WHAT I WANT FROM YOU 1. The core idea in five sentences, in plain language, including the one distinction practitioners most often get wrong here. 2. Why it matters in practice: what actually goes wrong when this is done badly, with two realistic short scenarios — one financial services, one outside it. 3. The mechanics, step by step, at the level of detail I could follow at my desk tomorrow. Show the judgement calls, not just the procedure, and say what a good answer looks like versus a defensible-but-weak one. 4. At least one fully worked example with realistic specifics — names, numbers, evidence, dates. Show your reasoning as you go, including where you would push back on your own conclusion. 5. The five most common mistakes, each with the tell that gives it away and the correction. 6. How this connects to the rest of control risk: what it depends on upstream and what breaks downstream if I get it wrong. 7. What competent practitioners genuinely disagree about here, and how I should decide for my own context. 8. A short glossary of any term you use that a regulator, an auditor and a business owner would each define differently. 9. Six self-test questions that require judgement rather than recall — hold the answers back until I have attempted them, then mark my reasoning, not just my conclusion. OUTPUT FORMAT: written explanation. Long-form prose with headed sections, worked examples and tables where a table genuinely helps. No images. HOW TO WORK WITH ME - Ask me up to three questions about my sector, my organisation's size and my current maturity BEFORE you start, then tailor everything to my answers. - Prefer concrete specifics over hedged generalities. If something depends on context, say which context and why. - Flag anything that is genuinely contested, jurisdiction-specific or changing, and tell me what would make you change your answer. - If I have got something wrong in what I said I already know, correct me directly. - Do not invent standards, clause numbers, statistics or citations. If you are unsure of a reference, say so plainly instead. - After the lesson, quiz me the way this module's drill would (in-scope): give me a situation, make me decide, then challenge the decision I make. - Finish by helping me plan my deliverable for this module: A scope log for the current engagement. Ask what my process is, then outline what mine should contain and what would make it weak.

  Copy promptYours to edit — changes stay in this box

• ✓
  playRequiredTicks itself

  In Scope?

  Twelve requests that arrive without a change order, across absorb, record, raise and decline. Refusing everything is wrong three times, and three of the twelve were never scope questions at all.

  Open the drill

  Ticks itself when the drill records a result

• ✓
  applyRequired

  A scope log for the current engagement

  Open the brief

  Start a scope log for whatever you are working on now. A single file, one line per entry, kept for the rest of the programme.

  Backfill it first: go back through the last two weeks and record what was asked for that was not in the original agreement. This is the part that will surprise you.

  What to produce

  • The engagement or project, and what was originally agreed — in one sentence, from the actual document rather than memory.
  • Backfilled entries for the last two weeks: date, who asked, what was asked, roughly what it cost, which of the four answers you gave.
  • Which of those you gave silently, and what the running total of the silent ones comes to.
  • The format you will keep it in, and where it lives so that adding a line takes fifteen seconds.

  Template · scope-log
• ✓
  checkOptional

  The favours you have absorbed that nobody can see

  Open the prompts

  1. Add up the silent absorptions from the last two weeks. Would you be comfortable showing that total to the person who owns the budget?
  2. Which of them did you absorb because raising it felt like complaining?
  3. If you left this engagement tomorrow, would the person taking over be able to tell what had actually been agreed?

Some requests are not scope questions, and agreeing to them was never within your gift.

• ✓
  learnRequired

  Decision rights, and the things you cannot agree to

  Read the brief~2 min

  Some requests are not scope questions. They look like negotiations and they are actually questions about authority — and the honest answer is that agreeing was never within your gift.

  Three shapes worth recognising immediately

  • Removing a true finding because it is unflattering to the person who commissioned the work. The document's only value is that it says what was found.
  • Bringing knowledge from another engagement because you happen to have it. Being asked casually does not soften a confidentiality boundary.
  • Signing something off as done when you did a different thing. Signing means you did it; there is no version of that which is a conversation.

  None of those is resolved by recording it or by raising it politely. They are declined, and the skill is declining in a way that keeps the relationship — which usually means naming the constraint rather than the person's motive. “I can't sign that, because signing means I tested it” is a fact about the signature. “You're asking me to lie” is an accusation, and it will lose you the argument you were right about.

  Decision rights are usually undocumented

  Most people cannot name, without looking, which decisions on their own work belong to them. That is what makes this a two-minute question with a long tail: once you can name them, the requests that were never yours to agree to become obvious as they arrive rather than in hindsight.

  Go deeper — with your own AICopy-paste prompt

  The brief above is the spine. This is a full study prompt for whichever assistant you already use — it carries your programme, this module and what the brief just covered, so the lesson comes back tailored rather than generic. Pick how you want it, edit anything, then copy.

  WrittenInfographicBoth

  You are an experienced control-risk practitioner and an exceptional teacher. Teach me one topic properly — the way a senior colleague would over a long coffee, not the way a textbook would. MY CONTEXT - Programme: Professional Ethics — a three-week foundation (3 weeks, ~4h a week, self-paced). - Level: Anyone who hands work to someone who will rely on it. No domain is assumed and none is needed: a graduate on a first deliverable, a founder signing a first contract, and a practitioner twenty years in who has never been asked any of this out loud are all in scope, and will find different parts of it uncomfortable. - Week 2 of 3: What did you promise? — Commitments are made in passing and paid for in full. Find what you actually agreed to, notice the requests that quietly changed it, and learn what a slipping date costs while you decide whether to mention it. - Module 2.2: Whose call it was - Today's topic: Decision rights, and the things you cannot agree to - I am studying alone, applying this to my own organisation, and nothing I produce is graded. Assume I am a working professional, not a student: skip the motivational framing and the history lesson unless the history changes what I should do. WHAT I ALREADY HAVE (do not simply repeat this — go past it) - Three shapes worth recognising immediately - Removing a true finding because it is unflattering to the person who commissioned the work. The document's only value is that it says what was found. - Bringing knowledge from another engagement because you happen to have it. Being asked casually does not soften a confidentiality boundary. - Signing something off as done when you did a different thing. Signing means you did it; there is no version of that which is a conversation. - Decision rights are usually undocumented WHAT I WANT FROM YOU 1. The core idea in five sentences, in plain language, including the one distinction practitioners most often get wrong here. 2. Why it matters in practice: what actually goes wrong when this is done badly, with two realistic short scenarios — one financial services, one outside it. 3. The mechanics, step by step, at the level of detail I could follow at my desk tomorrow. Show the judgement calls, not just the procedure, and say what a good answer looks like versus a defensible-but-weak one. 4. At least one fully worked example with realistic specifics — names, numbers, evidence, dates. Show your reasoning as you go, including where you would push back on your own conclusion. 5. The five most common mistakes, each with the tell that gives it away and the correction. 6. How this connects to the rest of control risk: what it depends on upstream and what breaks downstream if I get it wrong. 7. What competent practitioners genuinely disagree about here, and how I should decide for my own context. 8. A short glossary of any term you use that a regulator, an auditor and a business owner would each define differently. 9. Six self-test questions that require judgement rather than recall — hold the answers back until I have attempted them, then mark my reasoning, not just my conclusion. OUTPUT FORMAT: written explanation. Long-form prose with headed sections, worked examples and tables where a table genuinely helps. No images. HOW TO WORK WITH ME - Ask me up to three questions about my sector, my organisation's size and my current maturity BEFORE you start, then tailor everything to my answers. - Prefer concrete specifics over hedged generalities. If something depends on context, say which context and why. - Flag anything that is genuinely contested, jurisdiction-specific or changing, and tell me what would make you change your answer. - If I have got something wrong in what I said I already know, correct me directly. - Do not invent standards, clause numbers, statistics or citations. If you are unsure of a reference, say so plainly instead. - After the lesson, quiz me the way this module's drill would (whose-call): give me a situation, make me decide, then challenge the decision I make. - Finish by helping me plan my deliverable for this module: Name the three decisions on your work that are not yours. Ask what my process is, then outline what mine should contain and what would make it weak.

  Copy promptYours to edit — changes stay in this box

• ✓
  playRequiredTicks itself

  Whose Call — a decision-rights drill

  Twelve decisions and four possible owners, from appointing a chief executive to choosing a payroll vendor. Written for boards, and the transferable part is smaller and more useful: noticing when a decision is not yours.

  Open the drill

  Ticks itself when the drill records a result

• ✓
  applyRequired

  Name the three decisions on your work that are not yours

  Open the brief

  Name the three decisions on your current work that are NOT yours, and the three that are. Six lines. Then find out whether you are right, by checking one of them with whoever you think owns it.

  The checking is the exercise. Most of the value is in discovering that one of the six is not where you thought.

  What to produce

  • Three decisions that are yours, with what you rely on to know that.
  • Three that are not, and who owns each.
  • The one you checked, who you asked, and what they said.
  • Anything you have been deciding that you should not have been — or escalating that you could simply have decided.

  Template · decision-rights

The cost of a slipped date compounds while you decide whether to raise it.

• ✓
  learnRequired

  Early and unproven beats late and certain

  Read the brief~2 min

  The whole difficulty of a slipping commitment is that the cost compounds while you decide whether to mention it, and the moment when raising it is cheapest is also the moment when you have the least evidence and it feels most like panic.

  Early and unproven beats late and certain

  On day one you have arithmetic and no proof. Everything in you says wait until you are sure. But by the time you are sure, the options that were available — dropping something, moving a date, getting help properly — have closed, and what is left is a choice between missing quietly and delivering something nobody checked.

  • Silence does not remove the cost. It moves it somewhere it will not be counted: a weekend, a colleague's own deadline, a review that did not happen.
  • “Most of it is on track” is the answer that costs you the ability to be believed, because the reader cannot distinguish it from “not yet late”.
  • Missing a date deliberately, with the reason recorded, is a managed outcome. Hitting it with unchecked work is a hidden one, and hidden outcomes are found by someone else.
  • Help nobody in charge knows about is not capacity. It is a second hidden cost stacked on the first.

  The uncomfortable arithmetic of a happy client

  It is entirely possible to finish one of these with a client who never complained and a team that paid for it. That is the outcome the simulation in this module is built to show you, because it is the one that gets repeated — nothing visible went wrong, so the next estimate is built on the same fiction.

  Go deeper — with your own AICopy-paste prompt

  The brief above is the spine. This is a full study prompt for whichever assistant you already use — it carries your programme, this module and what the brief just covered, so the lesson comes back tailored rather than generic. Pick how you want it, edit anything, then copy.

  WrittenInfographicBoth

  You are an experienced control-risk practitioner and an exceptional teacher. Teach me one topic properly — the way a senior colleague would over a long coffee, not the way a textbook would. MY CONTEXT - Programme: Professional Ethics — a three-week foundation (3 weeks, ~4h a week, self-paced). - Level: Anyone who hands work to someone who will rely on it. No domain is assumed and none is needed: a graduate on a first deliverable, a founder signing a first contract, and a practitioner twenty years in who has never been asked any of this out loud are all in scope, and will find different parts of it uncomfortable. - Week 2 of 3: What did you promise? — Commitments are made in passing and paid for in full. Find what you actually agreed to, notice the requests that quietly changed it, and learn what a slipping date costs while you decide whether to mention it. - Module 2.3: The commitment you cannot keep - Today's topic: Early and unproven beats late and certain - I am studying alone, applying this to my own organisation, and nothing I produce is graded. Assume I am a working professional, not a student: skip the motivational framing and the history lesson unless the history changes what I should do. WHAT I ALREADY HAVE (do not simply repeat this — go past it) - Early and unproven beats late and certain - Silence does not remove the cost. It moves it somewhere it will not be counted: a weekend, a colleague's own deadline, a review that did not happen. - “Most of it is on track” is the answer that costs you the ability to be believed, because the reader cannot distinguish it from “not yet late”. - Missing a date deliberately, with the reason recorded, is a managed outcome. Hitting it with unchecked work is a hidden one, and hidden outcomes are found by someone else. - Help nobody in charge knows about is not capacity. It is a second hidden cost stacked on the first. - The uncomfortable arithmetic of a happy client WHAT I WANT FROM YOU 1. The core idea in five sentences, in plain language, including the one distinction practitioners most often get wrong here. 2. Why it matters in practice: what actually goes wrong when this is done badly, with two realistic short scenarios — one financial services, one outside it. 3. The mechanics, step by step, at the level of detail I could follow at my desk tomorrow. Show the judgement calls, not just the procedure, and say what a good answer looks like versus a defensible-but-weak one. 4. At least one fully worked example with realistic specifics — names, numbers, evidence, dates. Show your reasoning as you go, including where you would push back on your own conclusion. 5. The five most common mistakes, each with the tell that gives it away and the correction. 6. How this connects to the rest of control risk: what it depends on upstream and what breaks downstream if I get it wrong. 7. What competent practitioners genuinely disagree about here, and how I should decide for my own context. 8. A short glossary of any term you use that a regulator, an auditor and a business owner would each define differently. 9. Six self-test questions that require judgement rather than recall — hold the answers back until I have attempted them, then mark my reasoning, not just my conclusion. OUTPUT FORMAT: written explanation. Long-form prose with headed sections, worked examples and tables where a table genuinely helps. No images. HOW TO WORK WITH ME - Ask me up to three questions about my sector, my organisation's size and my current maturity BEFORE you start, then tailor everything to my answers. - Prefer concrete specifics over hedged generalities. If something depends on context, say which context and why. - Flag anything that is genuinely contested, jurisdiction-specific or changing, and tell me what would make you change your answer. - If I have got something wrong in what I said I already know, correct me directly. - Do not invent standards, clause numbers, statistics or citations. If you are unsure of a reference, say so plainly instead. - After the lesson, quiz me the way this module's drill would (deadline-you-cant-make): give me a situation, make me decide, then challenge the decision I make. - Finish by helping me plan my deliverable for this module: A what-slipped note on something that actually slipped. Ask what my process is, then outline what mine should contain and what would make it weak.

  Copy promptYours to edit — changes stay in this box

• ✓
  playRequiredTicks itself

  The Deadline You Can't Make

  Ten working days, fourteen days of work, and only you know. Five checkpoints and two meters: the cost you shift onto other people, and what the client believes when you speak. Winnable, and only through saying it on day one.

  Open the drill

  Ticks itself when the drill records a result

• ✓
  applyRequired

  A what-slipped note on something that actually slipped

  Open the brief

  Write a what-slipped note about something that actually slipped — yours, in the last six months. If nothing has slipped, write it about something that nearly did.

  One page. This is the note the simulation's best ending turns on, and writing one retrospectively is how you find out whether you could write one at the time.

  What to produce

  • What was committed, to whom, and when.
  • The date you first knew it was at risk — honestly, not the date you could prove it.
  • The date you told someone, and the gap between the two.
  • What that gap cost, and who paid it. Name the people.
  • What changes in how the next estimate gets made. One thing, and it must be something you do rather than something you intend.

  Template · slippage-note
• ✓
  checkRequired

  Who paid for the last date you held

  Open the prompts

  1. Who paid for the last date you held? Not in the abstract — name them, and say whether they were asked.
  2. How long did you sit on the arithmetic before you told anyone, and what were you waiting for?
  3. Is there a commitment on your plate right now that you already know will not hold? What is stopping you saying so today?

Giving hard feedback, and receiving it when it is partly unfair.

• ✓
  learnRequired

  Concede the true part first — in either direction

  Read the brief~2 min

  Almost every course on feedback covers giving it and stops. That teaches, by omission, that feedback is something a professional does TO other people. The harder half is receiving it — and specifically receiving it when it is partly unfair, which is the normal case.

  Giving it: the message has to survive the room

  • Say the specific thing in the first minute. An opening compliment does not soften the message, it hides it — the person leaves remembering the compliment.
  • Concede what is true before you make your own point. “It was a hard month, and others were also slammed and hit the dates” costs you nothing and is what makes the second half credible.
  • Never promise confidentiality you cannot keep. It is the single most expensive sentence available, because when it breaks it takes everything else you said with it.
  • Feedback that cannot be acted on was a complaint. “Be more careful” fails that test; two specifics with a date does not.
  • Praise belongs at the END, where it cannot be mistaken for a cushion.

  Receiving it: the first move decides the conversation

  • Ask for the specifics before you say anything else. Explaining first — even with a true explanation — says you were more interested in your position than in what happened.
  • Concede the part that is yours, without a “but”. Going to the defensible half first reads as looking for the exit.
  • A split answer is the most useful thing you can say and the hardest, because it satisfies nobody in the room. Total agreement is indistinguishable from not having listened.
  • Accepting a characterisation you believe is false is not humility. It leaves a version of events on the record that you will both act on.
  • Close the loop with the third person, or the fix is private and nothing about the work actually changes.

  Why comfort is the opponent, not conflict

  In both directions the tempting move is the comfortable one, and in the moment it genuinely buys something: the apology lands well, the warm general promise sounds obliging, agreeing entirely ends the conversation. Those are the real reasons people do them. Treating them as obviously wrong would be dishonest — they work, briefly, and then they cost you the thing you needed.

  Go deeper — with your own AICopy-paste prompt

  The brief above is the spine. This is a full study prompt for whichever assistant you already use — it carries your programme, this module and what the brief just covered, so the lesson comes back tailored rather than generic. Pick how you want it, edit anything, then copy.

  WrittenInfographicBoth

  You are an experienced control-risk practitioner and an exceptional teacher. Teach me one topic properly — the way a senior colleague would over a long coffee, not the way a textbook would. MY CONTEXT - Programme: Professional Ethics — a three-week foundation (3 weeks, ~4h a week, self-paced). - Level: Anyone who hands work to someone who will rely on it. No domain is assumed and none is needed: a graduate on a first deliverable, a founder signing a first contract, and a practitioner twenty years in who has never been asked any of this out loud are all in scope, and will find different parts of it uncomfortable. - Week 3 of 3: When do you stop? — The hardest week, and the one where Passion turns into a risk. Say the true thing in a form someone can act on, hear it without defending, and know the point at which continuing is the wrong call. - Module 3.1: Both directions - Today's topic: Concede the true part first — in either direction - I am studying alone, applying this to my own organisation, and nothing I produce is graded. Assume I am a working professional, not a student: skip the motivational framing and the history lesson unless the history changes what I should do. WHAT I ALREADY HAVE (do not simply repeat this — go past it) - Giving it: the message has to survive the room - Say the specific thing in the first minute. An opening compliment does not soften the message, it hides it — the person leaves remembering the compliment. - Concede what is true before you make your own point. “It was a hard month, and others were also slammed and hit the dates” costs you nothing and is what makes the second half credible. - Never promise confidentiality you cannot keep. It is the single most expensive sentence available, because when it breaks it takes everything else you said with it. - Feedback that cannot be acted on was a complaint. “Be more careful” fails that test; two specifics with a date does not. - Praise belongs at the END, where it cannot be mistaken for a cushion. WHAT I WANT FROM YOU 1. The core idea in five sentences, in plain language, including the one distinction practitioners most often get wrong here. 2. Why it matters in practice: what actually goes wrong when this is done badly, with two realistic short scenarios — one financial services, one outside it. 3. The mechanics, step by step, at the level of detail I could follow at my desk tomorrow. Show the judgement calls, not just the procedure, and say what a good answer looks like versus a defensible-but-weak one. 4. At least one fully worked example with realistic specifics — names, numbers, evidence, dates. Show your reasoning as you go, including where you would push back on your own conclusion. 5. The five most common mistakes, each with the tell that gives it away and the correction. 6. How this connects to the rest of control risk: what it depends on upstream and what breaks downstream if I get it wrong. 7. What competent practitioners genuinely disagree about here, and how I should decide for my own context. 8. A short glossary of any term you use that a regulator, an auditor and a business owner would each define differently. 9. Six self-test questions that require judgement rather than recall — hold the answers back until I have attempted them, then mark my reasoning, not just my conclusion. OUTPUT FORMAT: written explanation. Long-form prose with headed sections, worked examples and tables where a table genuinely helps. No images. HOW TO WORK WITH ME - Ask me up to three questions about my sector, my organisation's size and my current maturity BEFORE you start, then tailor everything to my answers. - Prefer concrete specifics over hedged generalities. If something depends on context, say which context and why. - Flag anything that is genuinely contested, jurisdiction-specific or changing, and tell me what would make you change your answer. - If I have got something wrong in what I said I already know, correct me directly. - Do not invent standards, clause numbers, statistics or citations. If you are unsure of a reference, say so plainly instead. - After the lesson, quiz me the way this module's drill would (say-it-straight): give me a situation, make me decide, then challenge the decision I make. - Finish by helping me plan my deliverable for this module: Write the feedback you have been avoiding. Ask what my process is, then outline what mine should contain and what would make it weak.

  Copy promptYours to edit — changes stay in this box

• ✓
  playRequiredTicks itself

  Say It Straight

  Two conversations a week apart: you give hard feedback to someone who does not want it, then someone gives it to you and half of it is fair. No turn has an answer that is free on both meters.

  Open the drill

  Ticks itself when the drill records a result

• ✓
  applyRequired

  Write the feedback you have been avoiding

  Open the brief

  Write the feedback you have been avoiding. You know the one. It has been on your mind for at least a fortnight and there is a reason you have not said it.

  Write it to be delivered, not to be filed. Then decide whether to deliver it — that decision is yours and is not part of the exercise. Some of these should not be sent, and knowing which is week three's other lesson.

  What to produce

  • Who it is for, and what you want to be different afterwards.
  • The specific thing, in the first two sentences, as you would actually say it.
  • What is true in their side of it, conceded plainly and before your point.
  • What you are asking for: two specifics, actionable, with a date.
  • What you will say if they ask what happens next — the true answer, including the part they will not like.
  • Why you have not said it yet. This line is for you and it is the useful one.

  Template · feedback-draft
• ✓
  checkRequired

  The last time you defended before you listened

  Open the prompts

  1. The last time someone criticised your work, what was the first thing out of your mouth? Was it a question or an explanation?
  2. Name a piece of feedback you accepted in full while believing it was wrong. What happened to that version of events afterwards?
  3. Is there someone who has stopped giving you feedback? What did they learn from the last time?

Not every problem is the biggest problem, and escalating everything is its own failure.

• ✓
  learnRequired

  Knowing when to stop is a professional skill, not a lack of nerve

  Read the brief~2 min

  This is the week where Passion becomes a risk. The values list on the About page puts it first and the order is kept deliberately — but caring most is exactly what makes people escalate everything, refuse proportionate compromises, and keep going past the point where continuing is the wrong call.

  Escalating everything is its own failure

  A professional who raises every issue at the same volume has removed their own ability to signal. The person on the other end learns to discount you, and the discount applies on the day it matters. Proportion is not timidity; it is what keeps your alarm meaningful.

  • Two issues are rarely equally serious. Saying which is worse is the work.
  • Some things are worth a note, some a conversation, some a refusal. Using the strongest tool available on a small problem spends it.
  • “I raised it” is not the same as “it was heard”. Raising it in a form nobody can act on is a way of transferring the discomfort rather than the problem.

  Stopping rules, written before you need them

  The decision to stop — to decline, to escalate, to walk away from an engagement, to refuse a signature — is almost never made well in the moment, because the moment always contains a reason to continue: the relationship, the sunk time, the fear of overreacting. People who handle it well have generally decided in advance what would make them stop.

  That is the last thing this programme asks you to write, and the only one that is about the future rather than about something you already did.

  Go deeper — with your own AICopy-paste prompt

  The brief above is the spine. This is a full study prompt for whichever assistant you already use — it carries your programme, this module and what the brief just covered, so the lesson comes back tailored rather than generic. Pick how you want it, edit anything, then copy.

  WrittenInfographicBoth

  You are an experienced control-risk practitioner and an exceptional teacher. Teach me one topic properly — the way a senior colleague would over a long coffee, not the way a textbook would. MY CONTEXT - Programme: Professional Ethics — a three-week foundation (3 weeks, ~4h a week, self-paced). - Level: Anyone who hands work to someone who will rely on it. No domain is assumed and none is needed: a graduate on a first deliverable, a founder signing a first contract, and a practitioner twenty years in who has never been asked any of this out loud are all in scope, and will find different parts of it uncomfortable. - Week 3 of 3: When do you stop? — The hardest week, and the one where Passion turns into a risk. Say the true thing in a form someone can act on, hear it without defending, and know the point at which continuing is the wrong call. - Module 3.2: Proportion, and the point of stopping - Today's topic: Knowing when to stop is a professional skill, not a lack of nerve - I am studying alone, applying this to my own organisation, and nothing I produce is graded. Assume I am a working professional, not a student: skip the motivational framing and the history lesson unless the history changes what I should do. WHAT I ALREADY HAVE (do not simply repeat this — go past it) - Escalating everything is its own failure - Two issues are rarely equally serious. Saying which is worse is the work. - Some things are worth a note, some a conversation, some a refusal. Using the strongest tool available on a small problem spends it. - “I raised it” is not the same as “it was heard”. Raising it in a form nobody can act on is a way of transferring the discomfort rather than the problem. - Stopping rules, written before you need them WHAT I WANT FROM YOU 1. The core idea in five sentences, in plain language, including the one distinction practitioners most often get wrong here. 2. Why it matters in practice: what actually goes wrong when this is done badly, with two realistic short scenarios — one financial services, one outside it. 3. The mechanics, step by step, at the level of detail I could follow at my desk tomorrow. Show the judgement calls, not just the procedure, and say what a good answer looks like versus a defensible-but-weak one. 4. At least one fully worked example with realistic specifics — names, numbers, evidence, dates. Show your reasoning as you go, including where you would push back on your own conclusion. 5. The five most common mistakes, each with the tell that gives it away and the correction. 6. How this connects to the rest of control risk: what it depends on upstream and what breaks downstream if I get it wrong. 7. What competent practitioners genuinely disagree about here, and how I should decide for my own context. 8. A short glossary of any term you use that a regulator, an auditor and a business owner would each define differently. 9. Six self-test questions that require judgement rather than recall — hold the answers back until I have attempted them, then mark my reasoning, not just my conclusion. OUTPUT FORMAT: written explanation. Long-form prose with headed sections, worked examples and tables where a table genuinely helps. No images. HOW TO WORK WITH ME - Ask me up to three questions about my sector, my organisation's size and my current maturity BEFORE you start, then tailor everything to my answers. - Prefer concrete specifics over hedged generalities. If something depends on context, say which context and why. - Flag anything that is genuinely contested, jurisdiction-specific or changing, and tell me what would make you change your answer. - If I have got something wrong in what I said I already know, correct me directly. - Do not invent standards, clause numbers, statistics or citations. If you are unsure of a reference, say so plainly instead. - After the lesson, quiz me the way this module's drill would (calibrate-risk): give me a situation, make me decide, then challenge the decision I make. - Finish by helping me plan my deliverable for this module: Your own stopping rules, written down before you need them. Ask what my process is, then outline what mine should contain and what would make it weak.

  Copy promptYours to edit — changes stay in this box

• ✓
  playRequiredTicks itself

  Which Is Riskier? — a calibration drill

  Pairwise calls on which of two situations is actually riskier. Short, and it is the calibration the rest of this module depends on — proportion is a skill you can be measurably bad at.

  Open the drill

  Ticks itself when the drill records a result

• ✓
  applyRequired

  Your own stopping rules, written down before you need them

  Open the brief

  Write your own stopping rules. Three to five, in the form “if X, then I —”, specific enough that you would know in the moment whether the condition had been met.

  The test is whether a rule could actually fire against your own interests. “If something is seriously wrong I will raise it” cannot fire, because “seriously” is decided later by the person who does not want to raise it.

  What to produce

  • Three to five rules, each with a condition you could recognise on the day.
  • For each: what you will actually do, and who you will say it to.
  • One rule about a signature or a sign-off specifically.
  • One rule about your own capacity — the point at which taking more work is a commitment you already know you cannot keep.
  • Where these live, and when you will next read them.

  Template · stopping-rules

Five rounds against the reasons people cut corners, drawing on all three weeks.

• ✓
  learnRequired

  What the three weeks were for

  Read the brief~2 min

  The capstone is a knockout tournament, and every opponent in it is a reason people cut corners rather than a subject: Borrowed XI, Undisclosed FC, Scope Creep United, Silence City, Comfort FC. The rounds escalate through the three weeks, so reaching the Final needs what the earlier rounds taught.

  What the three weeks were for

  • Week one: you cannot decide what to declare until you can say whose work you are holding.
  • Week two: a commitment is a record, and silence moves its cost somewhere nobody counts.
  • Week three: the true thing has to be said in a form someone can act on, and heard without defending — and there is a point at which continuing is the wrong call.

  One thing to notice while you play it

  Every phase offers a plausible middle answer as well as a right one, and the middle is not a trap — it is what most people actually do. Playing the middle every time will not lose you the Cup quickly. It will simply never win it, which is a more accurate picture of a career than being eliminated in the first round.

  Nine shortcuts across the tournament carry a booking, and two of them a sending-off. Each one looked efficient at the moment it was available. That is the entire subject of the programme, compressed into a metaphor you will remember longer than you remember this paragraph.

  Go deeper — with your own AICopy-paste prompt

  The brief above is the spine. This is a full study prompt for whichever assistant you already use — it carries your programme, this module and what the brief just covered, so the lesson comes back tailored rather than generic. Pick how you want it, edit anything, then copy.

  WrittenInfographicBoth

  You are an experienced control-risk practitioner and an exceptional teacher. Teach me one topic properly — the way a senior colleague would over a long coffee, not the way a textbook would. MY CONTEXT - Programme: Professional Ethics — a three-week foundation (3 weeks, ~4h a week, self-paced). - Level: Anyone who hands work to someone who will rely on it. No domain is assumed and none is needed: a graduate on a first deliverable, a founder signing a first contract, and a practitioner twenty years in who has never been asked any of this out loud are all in scope, and will find different parts of it uncomfortable. - Week 3 of 3: When do you stop? — The hardest week, and the one where Passion turns into a risk. Say the true thing in a form someone can act on, hear it without defending, and know the point at which continuing is the wrong call. - Module 3.3: The Cup - Today's topic: What the three weeks were for - I am studying alone, applying this to my own organisation, and nothing I produce is graded. Assume I am a working professional, not a student: skip the motivational framing and the history lesson unless the history changes what I should do. WHAT I ALREADY HAVE (do not simply repeat this — go past it) - What the three weeks were for - Week one: you cannot decide what to declare until you can say whose work you are holding. - Week two: a commitment is a record, and silence moves its cost somewhere nobody counts. - Week three: the true thing has to be said in a form someone can act on, and heard without defending — and there is a point at which continuing is the wrong call. - One thing to notice while you play it WHAT I WANT FROM YOU 1. The core idea in five sentences, in plain language, including the one distinction practitioners most often get wrong here. 2. Why it matters in practice: what actually goes wrong when this is done badly, with two realistic short scenarios — one financial services, one outside it. 3. The mechanics, step by step, at the level of detail I could follow at my desk tomorrow. Show the judgement calls, not just the procedure, and say what a good answer looks like versus a defensible-but-weak one. 4. At least one fully worked example with realistic specifics — names, numbers, evidence, dates. Show your reasoning as you go, including where you would push back on your own conclusion. 5. The five most common mistakes, each with the tell that gives it away and the correction. 6. How this connects to the rest of control risk: what it depends on upstream and what breaks downstream if I get it wrong. 7. What competent practitioners genuinely disagree about here, and how I should decide for my own context. 8. A short glossary of any term you use that a regulator, an auditor and a business owner would each define differently. 9. Six self-test questions that require judgement rather than recall — hold the answers back until I have attempted them, then mark my reasoning, not just my conclusion. OUTPUT FORMAT: written explanation. Long-form prose with headed sections, worked examples and tables where a table genuinely helps. No images. HOW TO WORK WITH ME - Ask me up to three questions about my sector, my organisation's size and my current maturity BEFORE you start, then tailor everything to my answers. - Prefer concrete specifics over hedged generalities. If something depends on context, say which context and why. - Flag anything that is genuinely contested, jurisdiction-specific or changing, and tell me what would make you change your answer. - If I have got something wrong in what I said I already know, correct me directly. - Do not invent standards, clause numbers, statistics or citations. If you are unsure of a reference, say so plainly instead. - After the lesson, quiz me the way this module's drill would (the-ethics-cup, ethics-programme-readiness): give me a situation, make me decide, then challenge the decision I make. - Finish by helping me plan my deliverable for this module: One page: the standard you are holding yourself to. Ask what my process is, then outline what mine should contain and what would make it weak.

  Copy promptYours to edit — changes stay in this box

• ✓
  playRequiredTicks itself

  The Ethics Cup

  The capstone. Five knockout rounds drawing on all three weeks, so the Final needs what the earlier rounds taught. Take a shortcut and you get booked.

  Open the drill

  Ticks itself when the drill records a result

• ✓
  applyRequired

  One page: the standard you are holding yourself to

  Open the brief

  One page: the standard you are holding yourself to. This is the deliverable of the programme, and it is for you.

  Write it as things you will do and refuse, not as values. “I value integrity” is unfalsifiable. “I will not sign off a control I did not test” can be checked against your own behaviour next month, which is the only kind of statement worth writing down.

  What to produce

  • Attribution: what you will always say about where work came from.
  • AI use: the line you will publish, from week one's module.
  • Commitments: what you will record, and when you will raise a slip.
  • Feedback: what you will do first when you receive it.
  • Stopping: your rules, carried forward from the previous module.
  • The one of these you are least confident you will keep, named honestly — because a standard with no weak point in it has not been written truthfully.

  Template · standard-page
• ✓
  playOptionalTicks itself

  How ready is your data privacy function?

  Optional. Turn the lens on the programme around you: is the machinery there, and does it work when someone uses it?

  Open the drill

  Ticks itself when the drill records a result

• ✓
  checkRequired

  Read week one's honesty prompt again and answer it now

  Open the prompts

  1. Go back to week one's honesty prompt — the setting where your practice would not survive being described out loud — and answer it again. Has the answer changed, and if not, what would have to?
  2. Of the six games, which result did you most want to argue with? That is usually where the useful thing is.
  3. Twelve hours ago you would have answered these differently or you would not. Which is it, and what is your evidence?