Library › Legal & Compliance AI › Compliance and risk
Reconcile claims with test results from the old and new product
- For Compliance and risk
- You practice Catching mistakes and made-up facts
- Time about 5 minutes
- Level Advanced
About this exercise
The product page describes the current v2 bottle, but some evidence in the notes covers v1. Write a prompt for a markdown table with columns Claim | Evidence for v2 | Status | Question, one row per claim C1 to C4, under 190 words. Status is only Supported, Not supported or Needs review. Use v2 evidence only; if there is none, say so. The last line must read: Open for reviewer: followed by the ids that are not Supported, in order.
Common mistakes when you ask AI for this
- Using the v1 result. AI cites the 24 hour v1 test as support for the v2 page. Try: Say: "Use v2 evidence only; ignore v1 results."
- Wrong status for pending. AI marks a pending certificate as Supported or Not supported. Try: Say: "Needs review when evidence is pending or partial."
- Missing the closing line. AI ends with its own summary or recommendation. Try: Say: "Last line exactly: Open for reviewer: then the ids not Supported."
- Trusting the AI to count words. AI is bad at counting and usually runs long, so a vague "keep it short" lands over the limit. Try: Cap each part instead of the total, for example "each line under 15 words".
What a strong prompt includes
- Restricts to v2
- Fixes the statuses
- Sets the table columns
- Closing line
- Caps cell length
Related exercises
- Accrue unpaid wages for the working days after the last pay period (Finance AI · Accountants)
- Age a returns queue in business days and flag the ones over the limit (Customer Support & Success AI · Customer support teams)
- Alert sales to a competitor's price change, facts only (Research & AI Search · Product managers)
- Annualize recurring revenue without the one-time fee (Data AI · Finance and accounting)
- Answer a business owner's contract questions or send to legal (Legal & Compliance AI · Legal teams)