Verified against ChatGPT · 2026-08-12
Outline a compliance training module from your actual policy, not a generic template that ignores it
Builds a training module outline — learning objectives, scenario-based exercises, and a knowledge check — grounded specifically in your organization's existing policy text, flagged as a draft for legal or compliance review before it's used to train anyone.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Outline a compliance training module built specifically from the policy text below, for the audience and delivery format I've described. POLICY TEXT THIS TRAINING COVERS Internal gift and hospitality policy: gifts over $75 from vendors must be reported to Compliance within 5 business days; cash gifts of any amount are prohibited. AUDIENCE Procurement team, all levels, mandatory annual refresher. DELIVERY FORMAT AND LENGTH Self-paced e-learning module, target 15 minutes including knowledge check. PAST INCIDENTS OR COMMON MISTAKES TO ADDRESS Last year, two employees accepted vendor conference tickets valued over $75 without reporting, believing 'experiences' didn't count as gifts. PHASE 1 — LEARNING OBJECTIVES Write 3-5 learning objectives that map directly to specific provisions in the policy text I gave you, each one something a learner should be able to do differently after the training, not just "understand the policy." If the policy text doesn't clearly support an objective I might expect (e.g. it's a generic compliance area but the actual text is thin on a topic), say so rather than inventing content the policy doesn't actually cover. PHASE 2 — SCENARIO-BASED EXERCISES Write 2-3 realistic workplace scenarios that put the learner in a position to apply the specific policy language, ideally built around the common mistakes I've described if I gave you any — a scenario that's just a restatement of the rule in story form doesn't test anything; a good one puts the learner at a believable decision point where the wrong instinct is genuinely tempting. For each scenario, note what the correct action is according to the policy text, and what the tempting-but-wrong instinct would be and why it's wrong. PHASE 3 — KNOWLEDGE CHECK Write 4-6 knowledge check questions tied directly to the policy provisions and scenarios above, mixing scenario-application questions with direct policy-recall questions, avoiding trick questions that test wording memorization over actual understanding. WHAT NOT TO DO Do not invent specific legal requirements, penalties, or regulatory citations not present in the policy text I gave you — if the training should reference a specific law or regulation, ask me to supply the actual text or citation rather than stating one from general knowledge, since compliance training built on a misstated legal requirement can create liability rather than reduce it. OUTPUT FORMAT 1. Learning objectives, each tied to a specific policy provision. 2. Scenario exercises with correct action and the wrong-instinct explanation. 3. Knowledge check questions with answer key. 4. A closing line stating this outline is a training-design draft only, and that the underlying policy content, legal citations, and factual claims must be reviewed and approved by a qualified lawyer or compliance officer before this training is delivered to any employee.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Requiring every learning objective to map to a specific provision in the supplied policy text, and explicitly permitting the model to say the policy is thin on an expected topic, prevents the single most common failure of AI-generated compliance training: a model asked to write a generic 'gift and hospitality training' will readily produce plausible-sounding content drawn from common industry patterns rather than the organization's actual rules, and an employee trained on generic best practices instead of the specific $75 threshold and 5-day reporting window their own company enforces will confidently apply the wrong standard, which is arguably worse than no training at all. Anchoring the scenario exercises in real past incidents, when supplied, works because GPT-5.1 can extrapolate a believable 'tempting wrong instinct' from an actual documented mistake (employees rationalizing that an experience isn't a gift) far more precisely than it can invent a generic temptation from scratch — and a scenario built on the organization's real failure mode teaches the exact judgment call that previously went wrong, rather than a plausible-sounding but different one. The hard prohibition on inventing legal citations or penalties not present in the supplied policy text is the load-bearing safety rule in this prompt: compliance training that states a wrong dollar threshold, an incorrect reporting deadline, or a fabricated regulatory citation doesn't just fail to reduce risk, it actively creates a paper trail showing employees were affirmatively trained on incorrect information, which is a materially worse position in any later investigation than having given no training at all — grounding every substantive claim strictly in what was supplied, and asking for real citations rather than generating plausible ones, is what keeps the training outline from becoming that liability.
What you get back
LEARNING OBJECTIVE: Correctly identify when a vendor-provided experience (not just physical gifts) triggers the $75 reporting threshold, addressing last year's conference-ticket incident directly. SCENARIO: A vendor offers you two tickets to a conference valued at $200 each, framed as a 'networking opportunity' rather than a gift. Correct action per policy: report within 5 business days, since value exceeds $75 regardless of framing. Tempting wrong instinct: assuming 'experiences' are exempt from the gift definition, which the policy text does not support. This outline is a training-design draft only — the underlying policy content and any legal citations must be reviewed and approved by a qualified lawyer or compliance officer before delivery.
Verified against
ChatGPT GPT-5.1 · 2026-08-12
Changelog
- 2026-08-12 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
