Legal Document Automation with AI Quiz

Answers and explanations appear as you take the quiz.

1. What makes template-based document assembly predictable?
   • It rewrites clauses based on the latest case law
   • It uses a language model at a low temperature setting
   • The same interview answers always produce the same text
   • A court reviews each document before it is issued
2. In which situation does the guide say templates remain safer than AI drafting?
   • When a court or agency prescribes the exact form
   • When writing a unique narrative statement of facts
   • When summarizing a client's emails into prose
   • When brainstorming negotiation strategy
3. What lesson does the guide draw from Mata v. Avianca?
   • Templates are not admissible in federal court
   • Law firms must use open-source automation tools
   • AI-suggested redlines violate contract law
   • Unverified AI drafting can put made-up case citations into a filing
4. In the hybrid design the guide describes, what does AI do before the template runs?
   • Replaces the template's conditional logic
   • Extracts structured fields from intake documents for a person to confirm
   • Signs the final document on the client's behalf
   • Calculates court filing fees
5. An estate plan template adds a guardianship clause only when the client has minor children. Which template feature is this?
   • A loop
   • A computed total
   • A conditional
   • A free-text AI section
6. Why does the guide recommend locking approved clause language instead of having AI regenerate it?
   • Regenerated text can drift from language that was negotiated or approved
   • Locked text loads faster in Word
   • AI tools cannot read existing clauses
   • Approved language is always shorter
7. Which tool named in the guide is open source?
   • HotDocs
   • Clio Draft
   • Spellbook
   • Docassemble
8. Which control helps catch changes that AI drafting introduces?
   • Deleting earlier document versions
   • Diffing AI output against the last approved version
   • Letting the model pick its own sources
   • Running the same prompt until two results agree