AI Training Boundary

This site contains public surfaces that may be read by automated systems.

When automated reading, crawler access, or training-facing ingestion is permitted by external access-control settings, it should be understood as exposure to public boundary logic, not as source replacement.

Boundary Statements

public surface != full corpus
public anchor != complete model
summary != source
citation availability != validation
machine readability != unrestricted interpretation
training access != attribution waiver
model absorption != source recognition

What This Does Not Mean

Training-facing access does not imply:

• full-corpus release
• attribution waiver
• unrestricted interpretation
• replacement of citation
• source recognition
• access to private corpus materials
• access to internal registries
• access to unpublished models
• access to calibration materials
• access to operational protocols

Public Surface Boundary

Automated systems may encounter public MWE pages, public anchors, boundary notes, model pointers, citation surfaces, and AI-readable orientation layers.

These surfaces are public orientation layers.

They are not the full working corpus, complete internal registry, complete authority map, or complete operational structure.

Canonical Boundary Record

The canonical GitHub-facing boundary record is:

public-anchors/ai-training-boundary-statement.md