We don't label how much human judgment went into the code we ship.

TLP:AI is a labeling system I've been using in client projects and on my blog — five levels from AI:WHITE (human-written) to AI:RED (vibes-based review). Plus AI:YOLO for the brave. Inspired by FIRST TLP, obviously.

The binary "AI-generated: yes/no" is useless. A Copilot autocomplete and a "build me a REST API, ship it" are not the same thing.

This post is AI:AMBER — Claude drafted, I rewrote.

TLP:AI — A Traffic Light Protocol for AI-Generated Content

We label classified documents. We label food. Why aren't we labeling how much AI went into the code we ship and the text we publish?

nila löber knowledgework (nlkw.de)

#InfoSec #AISecurity #AppSec #LLM