Constitutional AI and Rule Overrides
Anthropic's January 22, 2026 Claude Constitution runs 79 pages and is CC0. It moves from rule-based to reason-based alignment and establishes a four-tier priority hierarchy: (1) safety and supporting human oversight, (2) ethics, (3) Anthropic guidelines, (4) helpfulness. Behaviours split into hardcoded prohibitions (bioweapons uplift, CSAM) that operators and users cannot override and soft-coded defaults that operators can adjust within defined bounds. The 2022 original (Bai et al.) trained harmlessness via self-critique and RLAIF against a constitution. The honest caveat: reason-based alignment relies on the model generalising principles to unanticipated situations. Anthropic's own 2023 participatory experiment showed 50% divergence between public-sourced and corporate principles; the 2026 version did not incorporate those findings. A fielded agent sees inputs that its designers never saw. No rule list is long enough to cover them. No rule list is short enough to apply quickly under compute pressure. The practical question: how do you align an agent to principles that survive both a long tail of cases and fast inference? Rule-based alignment (RBA): list every disallowed thing. Fast to check, easy to audit, impossible to keep current, often over-refuses on close analogs it didn't anticipate. Reason-based alignment (the 2026 Claude Constitution): encode principles, let the model reason. Scales across unseen cases, harder to audit, failure mode is principle-misapplication rather than miss-the-rule. The 2026 Constitution takes…
Constitutional AI and Rule Overrides: Anthropic's January 22, 2026 Claude Constitution runs 79 pages and is CC0. It moves from rule-based to reason-based…
This free lesson is part of the AI Engineering from Scratch curriculum. Read the full explanation, run the lesson code, and verify the result in the interactive reader or from the repository source.
Browse the complete course catalog or open this lesson on GitHub.