What is Constitutional AI (CAI), the technique Anthropic uses in Claude's training?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Constitutional AI is like giving the model a written set of values and asking it to grade its own homework against those values — then training on the improved versions.
Full explanation below image
Full Explanation
Constitutional AI is Anthropic's training methodology where a 'constitution' — a set of principles and values — guides the model to critique and revise its own outputs. During training, Claude generates responses, then critiques them against the constitutional principles, then revises them. This RLAIF (Reinforcement Learning from AI Feedback) approach reduces reliance on human labelers for every example. Option A is wrong — CAI is a technical training method, not a legal framework. Option C describes encryption, unrelated. Option D describes hardware, also unrelated.