Library › Legal & Compliance AI › Compliance and risk

Recommend routing for an incident using the escalation criteria

  • For Compliance and risk
  • You practice Catching mistakes and made-up facts
  • Time about 5 minutes
  • Level Advanced

About this exercise

Apply the escalation criteria to the incident. Plain text, in this order: a line Facts: followed by exactly 3 dash lines; three lines (a), (b) and (c), each "Met" or "Not met" with a reason under 15 words; one line that is exactly one of: Recommendation: Escalate to Legal / Recommendation: Keep with Compliance / Recommendation: Need more information; and a last line exactly: The Compliance Officer decides. Use the latest figures. Do not say what any law requires.

Common mistakes when you ask AI for this

  • Using the first count. The AI uses 80 and decides (a) is not met. Try: Say: "Use the latest figures from the correction."
  • Treating Director as executive. The AI marks (c) as met. Try: Say: "Executive means VP level or above, as defined."
  • Free-text recommendation. The AI writes a paragraph of advice. Try: Say: "One line, exactly one of the three options."
  • Loose structure. The AI skips the Facts list or the closing line. Try: Say: "Facts: with exactly 3 dash lines, then (a), (b), (c)."
  • Trusting the AI to count words. AI is bad at counting. Ask for a total word limit and it often lands over it. Try: Cap each part instead, for example: "each line under 15 words."
  • Markdown. The AI bolds words or adds headings. Try: Say: "Plain text, no bold, no headings."

What a strong prompt includes

  • Facts list
  • Criteria lines
  • Latest figures
  • Exact recommendation
  • Exact last line

Practice this now ↑

Related exercises