Skip to content
Longterm Wiki
Index
Division·26NE1AhBX9·Record·Parent

Constitutional AI

Verdictunchecked0%
1 check · 4/13/2026

Reset by flagship-curate before re-verification

Our claim

entire record
Parent Org
Anthropic
Name
Constitutional AI
Division Type
team
Status
active
Start Date
December 2022
Notes
Training AI systems to follow principles through self-critique and RLAIF. Core alignment technique used in all Claude models.

Source evidence

1 src · 1 check
unverifiable85%Haiku 4.5 · 4/13/2026

NoteThe source text describes Constitutional AI as a research methodology and training approach, not as a formal organizational division or team unit. While the paper lists many Anthropic researchers as authors, it does not explicitly state that 'Constitutional AI' is an active division or team within Anthropic's organizational structure. The record claims Constitutional AI is a 'division' with 'type: team' and 'status: active', but the source only establishes it as a research method/paper. The organizational structure claim cannot be verified from this academic paper excerpt, which focuses on the technical methodology rather than organizational hierarchy.

Case № 26NE1AhBX9Filed 4/13/2026Confidence 0%