This document specifies the LLMin8 Measurement Protocol v1.0 --a versioned, reproducible framework for measuring AI visibility across large language models including ChatGPT, Claude, Gemini, and Perplexity. The protocol introduces: • Immutable protocol versioning• Cryptographic chain-of-custody (SHA-256 hashing of prompts and outputs)• Replicate-based agreement analysis• Bootstrap confidence intervals• A formal Confidence Tier system governing permissible claims• Scoring algorithm version control with in-place re-scoring prohibition AI visibility is defined as an operationalized construct measuring observable model outputs (mention, citation, rank position) under controlled prompt conditions. This protocol is proposed as a candidate measurement standard for AI visibility systems.
Lubna Rizwana Noor (Sun,) studied this question.