Library › Legal & Compliance AI › Compliance and risk

Flag the three campaign claims the notes contradict

  • For Compliance and risk
  • You practice Setting limits and rules
  • Time about 5 minutes
  • Level Beginner

About this exercise

Marketing sent six claims for an email campaign, with notes from the product team. Write a prompt that flags only the three claims the notes contradict or don't match, in the order they appear in the email. Ask for exactly three bullets, each the claim in quotes, a dash, then what the notes say instead in under 15 words. Plain text, no intro, and no comment on whether any claim is lawful.

Common mistakes when you ask AI for this

  • Flagging every claim. AI lists all six with a comment. Try: Say: "Flag only the claims the notes contradict or don't match, exactly three."
  • Vague reasons. AI says 'not fully supported' without the note's facts. Try: Say: "Give what the notes say instead, with the figure."
  • Legal verdicts. AI calls claims misleading or illegal. Try: Say: "Do not say whether any claim is lawful, false or compliant."
  • Too many bullets. AI lists everything it finds instead of stopping at your limit. Try: Say "at most N bullets, nothing else" and name the limit.
  • Getting markdown back. AI adds bold, headings and bullets by default, which show up as stray symbols when pasted. Try: Say "plain text, no bold, no headings, no asterisks".

What a strong prompt includes

  • Sets bullet count
  • Gives line format
  • Keeps email order
  • Bans legal verdicts
  • Plain text

Practice this now ↑

Related exercises