About validator.es

When AI participates in the creation of a document, the conditions of that participation are invisible, unverifiable, and therefore unaccountable. Validator.es exists to change that.

The problem

AI-assisted documents have become indistinguishable on the surface from fully human-authored ones. Detection tools respond by attempting to identify traces of the generative process. But detection assumes adversarial intent and fails systematically against honest, supervised use. Verifiable declaration is a different answer: whoever uses AI declares exactly how they did so, and that declaration is cryptographically anchored to the document.

The approach

The Canal Transparency Protocol (PTC) defines a declaration and verification standard for human–AI co-creation. It does not assess quality or decide whether AI use is appropriate. It records what happened, under what conditions, with what level of human supervision and for what purpose. Three complementary hashes make separate changes to the content, declaration or declared conditions detectable.

The origin

Validator.es is part of the ACS — El Canal ecosystem, a documentary corpus on human-AI co-creation built since 2023 and published openly with persistent identifiers. The tool implements the PTC corpus (R-01 v1.1 / PTC-F v1.2 / ARCH v0.3), whose conceptual predecessor is ACS-6. It was developed by Manuel Barrera Anglada (ORCID: 0009-0000-8891-7021), civil servant and president of the Association for Transparency for Citizens, as a practical application of a three-year methodological corpus.

The ecosystem

The corpus is published on Zenodo under CC BY 4.0 and indexed in OpenAIRE. The full conceptual framework — ACS, LEA, CAP, PTC — is accessible at elcanal.es.

In the 2026 regulatory context

From 2 August 2026, certain transparency obligations under the EU AI Act apply. In Spain, Organic Bill 121/000096 remains before Parliament as of 21 July 2026 and is not yet law. Validator does not replace visible notices, technical marking or other applicable obligations. It adds a distinct, complementary layer: a verifiable declaration of how AI participated, what human review occurred and which exact document version the declaration refers to.