AnyoneLearnAI
LearnPathsPracticeReferenceTools
Start

Reference

HubGlossaryHow-toCheatsheetsSnippetsAPIQuick starts

Tutorials · Paths

Reference · Glossary

Constitutional AI

Train models with written principles — critique and revise harmful outputs.

On this page

  • When to use
  • When not to
  • Example

#When to use

When designing, debugging, or explaining AI systems that mention this term in docs or code.

#When not to

When a simpler term (prompt, chat, search) is enough for your audience.

#Example

See the linked lesson for an interactive walkthrough.

Learn by doing → Learn by doing →

Related terms

  • Alignment
  • Safety filter
← All terms
AnyoneLearnAI · Tutorials · Paths · Capstone · Practice · AI Tools · Reference · About
HomeLearnPractice