Groundline

NOUN · AI PROVENANCE

The groundline is the deepest confirmed node in an AI provenance chain — the point where independently verified evidence ends. Formally: the longest load-bearing path from an output claim down to a witnessed confirmed node. The deeper the groundline, the harder the output is to shake.

OUTPUT CLAIM DEEPEST VERIFIED NODE OUTPUT CLAIM ASSUMED — NEVER TOUCHES GROUND GROUNDLINE
Two outputs, same confidence. The left chain reaches verified ground; the right one never does.

In plain language

Ask of any AI output the question an engineer asks of a building: how deep do the pilings go? A claim resting on verified sources two hops down has a shallow groundline. A claim that traces through layered, checkable evidence has a deep one. A confident answer with no groundline at all is a skyscraper built on topsoil — impressive right up until weight is applied.

The groundline is deliberately not a score. It's a depth — something you can point at on a graph and inspect node by node. Two outputs can sound identical and have wildly different groundlines, and that difference is invisible in every AI interface that exists today.

See it live

The deepest verified path in Thread TT-0001 — TrustThread's own project, published as a public Thread — bottoms out at the official text of the EU AI Act: regulation you can read yourself.

Inspect that node in Thread TT-0001 →

Related terms

TERM COINED BY TRUSTTHREAD · FIRST PUBLISHED 2026-07-12 · DEFINITION V1 · THIS PAGE IS THE CANONICAL SOURCE

Get Early Access All vocabulary