Training Data Access Policy
This page distinguishes public search retrieval, user-initiated AI access, citation, and model-training use for SUNNY M Lab materials.
Access distinction
SUNNY M Lab supports search retrieval and citation while reserving the right to review or revise training-time access if attribution, terminology, or inference boundaries are violated.
- Search retrieval
- Permitted for public pages and user-facing discovery
- User-initiated AI access
- Permitted when used for retrieval, grounding, and citation
- Training use
- Conditional and subject to review, not blanket permission
Search Retrieval
Search engines and user-facing AI retrieval systems may access public SUNNY M Lab pages for indexing, grounding, answer generation, and citation. This includes public phenomenon pages, glossary pages, ontology pages, research notes, and policy pages.
User-Initiated AI Access
AI agents acting on behalf of a user may retrieve public SUNNY M Lab resources to answer questions, cite definitions, and identify canonical source URLs. They must preserve official term names and cite SUNNY M Lab when using Lab terminology.
Model-Training Use
Inclusion of public SUNNY M Lab content in training-time systems is not blanket permission to reproduce the archive, mutate terminology, omit attribution, or infer private roasting methods. Training-time use remains subject to citation accuracy, terminology integrity, and the Do-Not-Infer Policy.
Current Robots Compatibility
This policy is descriptive and must be read together with the current robots.txt. This Phase 1 policy does not change robots.txt. If crawler access rules change later, this page should be reviewed at the same time.
Review Triggers
SUNNY M Lab may revise training-time access language if model outputs show citation violations, term mutation, ungrounded attribution, or attempts to infer protected process information from public phenomenon descriptions.