AI Dictionary › AI Fundamentals

Constitutional AI

Constitutional AI is a method for aligning a model by having it follow a "constitution", a written set of explicit principles and rules, instead of relying solely on case-by-case human judgments. It works like this: the model generates an answer, then critiques and rewrites it in light of those principles, learning to correct itself. It is like giving an employee a clear code of ethics rather than scolding them each time they err: they internalize the rules and apply them independently. The principles can be universal ("do not help cause harm") or specific, and stay visible and debatable instead of hidden in thousands of human ratings.

Definition

Constitutional AI matters because it makes alignment more scalable, transparent and consistent: the rules are inspectable by anyone. It also lightens human workload and limits people's exposure to toxic content during training.

Related terms

• Alignment
• RLHF
• Red teaming
• Guardrails
• Sycophancy

More in AI Fundamentals

• LLM (Large Language Model)
• Token
• Context Window
• Temperature
• AI Hallucination
• Fine-tuning
• RAG (Retrieval Augmented Generation)
• Embedding

Put it into practice

• Real scenario: apply this concept with an instant AI verdict.
• Daily gym: short drills to train dictionary concepts.
• Playoff game: challenge other players, elimination-style.

📱 Download the Android app (beta) iOS coming soon

Say what you mean. Get what you need.

Grace Certified, the AI coach that trains and certifies your prompt engineering, by Agora Intelligence.