Glossary · Verifiable AI

Authenticity

真正性

Data or content that is genuine, unaltered since creation, and of verified origin (provenance). It is the broader concept of integrity plus verification of where something came from.

Definition

Authenticity is the state in which you can trace who made what, when, and how it was handled afterward. Where integrity guarantees only that contents are unchanged, authenticity also covers the genuineness of the origin itself — whether it came from a legitimate source and is not an impersonation.

In Japanese e-records it is defined as one of three requirements — authenticity, legibility, and preservation — under the Electronic Books Preservation Act and medical-information-system guidelines. It is upheld through electronic signatures, timestamps, and management of operation history (provenance).

Lemma implementation

Lemma verifies provenance cryptographically without handing over the source data, proving authenticity. It fixes who created which version, when, and whether it was altered as a tamper-proof proof, and uses zero-knowledge proofs and selective disclosure to reveal only the attributes needed.

What authenticity means by domain (regulation & e-records / media & content / finance & DeFi / AI output) and how to verify it are gathered in the explainer hub, “What is authenticity”.

Get started

Build authenticity into the decisions that matter.