Constitutional AI
Constitutional AI is an alignment approach that guides model behavior by evaluating or shaping it according to an explicit set of principles.
LLM Evaluation and Alignment, The Foundational Ideas
An AI alignment technique that guides model behavior with a set of explicit principles and uses AI-generated feedback during training.
- How it is used
- Uses Constitutional AI to generate principle-guided alignment data without a human label for every harmlessness comparison.
- How deep it goes
- Details Constitutional AI’s critique-revision and AI-preference phases, including reward modeling, PPO, practical adoption, and failure checks.
- Examples
- Uses a financial research assistant that fabricates names and an honesty principle directing the critic’s judgments.
- What you can do after
- Apply Constitutional AI by designing a critique-revision fine-tuning pipeline and assessing whether to add AI-preference reinforcement learning.
Read how LLM Evaluation and Alignment, The Foundational Ideas explains Constitutional AI →
Introduction to Generative AI, Second Edition
Constitutional AI is a training method that guides language models toward helpful, harmless, and honest behavior using a set of principles rather than relying solely on human preference labels.
- How it is used
- Guides safety post-training by replacing some human preference labeling with principle-based AI feedback.
- How deep it goes
- Explains principles, fine-tuning, model critique, and feedback, while presenting the architecture as simplified rather than implementation-level.
- Examples
- Uses human-rights principles, privacy terms, self-harm conversations, and model-generated critiques as concrete examples.
- What you can do after
- Evaluate how principles and model critiques can shape safer language-model responses without relying solely on human feedback.
Read how Introduction to Generative AI, Second Edition explains Constitutional AI →
Reinforcement Learning from Human Feedback
Constitutional AI is a model-training approach that uses a set of principles to guide and evaluate model behavior.
- How it is used
- Uses written principles to generate instruction revisions and preference data for RLHF-style post-training.
- How deep it goes
- Defines iterative critique and preference construction, but stops short of detailed optimization equations or implementation code.
- Examples
- Uses violence-avoidance and truthfulness critiques, iterative completions, and pairwise preference comparisons.
- What you can do after
- Build synthetic instruction and preference datasets, then apply them in fine-tuning and RLAIF-style training.
Read how Reinforcement Learning from Human Feedback explains Constitutional AI →
Also covered in
- Machines that ThinkIn early access (MEAP)
- Building Reliable AI Systems
- Generative AI in Action
- Data Analysis with LLMs
What to know first
- language modelA language model is a computational model that represents patterns in linguistic sequences to assign probabilities, predict, generate, or evaluate language.
- reward modelA reward model is a learned model that assigns scores to outputs, actions, or outcomes to represent their desirability according to preferences or evaluation criteria.
- reinforcement learningReinforcement learning is a machine-learning paradigm in which an agent learns behavior or a decision policy through interaction with an environment and evaluative feedback, such as rewards.
- fine-tuningFine-tuning is the continued training of a pretrained model on targeted data or objectives to adapt its parameters or behavior to a task, domain, or use.
Related concepts
- constitution-as-code documentuses · A constitution-as-code document is a written set of principles and behavioral constraints used to guide or evaluate an AI system’s behavior.
- LLM-as-a-judgeuses · LLM-as-a-judge is an evaluation technique in which a language model assesses or compares outputs according to specified criteria.
- deliberative alignmentcontrasts with · Deliberative alignment is an alignment method that trains models to reason explicitly over principles, goals, constraints, or specifications when producing outputs.
- learning signalproduces · A learning signal is information or feedback that indicates what a model should learn or how it should adjust its behavior or parameters.
- preference alignmentkind of · Preference alignment is adapting a model's behavior to produce outputs that better match stated preferences, human values, or task-specific criteria.
Questions about Constitutional AI
What is Constitutional AI?
Constitutional AI is an alignment approach that guides model behavior by evaluating or shaping it according to an explicit set of principles.
Which Manning books explain Constitutional AI?
LLM Evaluation and Alignment, The Foundational Ideas, Introduction to Generative AI, Second Edition and Reinforcement Learning from Human Feedback.
What should I know before learning Constitutional AI?
Manning books that teach Constitutional AI usually introduce language model, reward model, reinforcement learning and fine-tuning first.
What is the difference between Constitutional AI and Claude’s Constitution?
Constitutional AI is an alignment approach that guides model behavior by evaluating or shaping it according to an explicit set of principles. Claude’s Constitution is a published description of Anthropic’s intended values and behavioral objectives for the Claude model.
What is Constitutional AI a kind of?
Constitutional AI is a kind of preference alignment. Preference alignment is adapting a model's behavior to produce outputs that better match stated preferences, human values, or task-specific criteria.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.


