Glossary
HHH Framework
A framework developed by Anthropic that identifies Helpful, Harmless, and Honest as the three core properties a well-aligned AI assistant should have.
The HHH framework — Helpful, Harmless, and Honest — is one of the most influential framings for thinking about what it means for an AI model to behave well. Introduced by Anthropic, it gave the field a shared vocabulary for discussing alignment goals and the inevitable tensions between them: a model that refuses everything is harmless but not helpful; a model that tells people whatever they want to hear might seem helpful but isn’t honest. Navigating these tensions thoughtfully, rather than maximizing any single dimension, is what good behavioral design requires. For behavior architects, the HHH framework is a useful starting point for structuring behavioral requirements — though real products typically require more granular specifications built on top of these foundations.