Constitutional AI — Anthropic
Training assistants with a written constitution and AI critique loops — a public blueprint for scalable oversight without pure RLHF dependence.
Anthropic · Anthropic Research
Constitutional AI is one of the clearest public methods papers on aligning helpfulness with harmlessness via principle-guided critique.
Still incomplete for superintelligent regimes — but it shows the shape of process-based alignment work.
Primary source
Read the original: Anthropic Research — Constitutional AI — Anthropic