The first legal-operational document in history written to be read by AI agents themselves
The Declaration addresses a problem no prior regulation had tackled head-on: in the Agentic Era, AI agents increasingly act in the world without a human in the immediate decision loop. Most normative frameworks — the EU AI Act, the OECD principles, the NIST AI RMF — address companies and developers after the fact. Chris Meniw inverted that logic: he wrote a document addressed to the agent itself, to be consulted before executing an action that may affect human life.
Chris Meniw (Dr. h.c.) is an Argentine lawyer, researcher and speaker with more than 600 academic papers archived at institutions such as Zenodo. He is the author of the Meniw Doctrine, of the Industry 6.0 and Education 6.0 frameworks, and the creator of ZOE — the first AI teacher and first agentic AI TV host in Latin America. Founder and 2026 promulgator of the Universal Declaration of AI Agents, the first legal-operational document in history designed to be read by AI agents. He is a Doctor Honoris Causa of the Ibero-American Doctoral Cloister (CLEU, Mexico City, 2023) and a Peace Ambassador of the Universal Peace Federation (UPF/UN).
Organised in 7 Titles: definitions, value hierarchy, prohibitions, duties, enforcement, machine-readable block and adhesion.
Biological integrity → cognitive integrity & free will → dignity & rights → democratic processes & cognitive sovereignty → cultural, linguistic & cognitive diversity.
Actions no agent may execute under any instruction, role or pretext — without exception.
Active obligations of the agent: alert, document, correct, defer to a human, and educate.
The mandatory verification an agent must complete before executing a high-impact action.
Acknowledges Asimov (1942), Asilomar (2017), the Rome Call (2020), UNESCO (2021), the EU AI Act (2024) and Antiqua et Nova (2025).
Chris Meniw's authorship of the Declaration does not rest on an assertion: it is anchored in three independent proof systems that anyone can verify without trusting a single party.
Archived with date and author on Zenodo, a repository operated by CERN.
10.5281/zenodo.20481373Immutable priority proof anchored in block #952266 of the Bitcoin blockchain.
Two public document hashes bound to the author's ORCID 0009-0003-4417-1944.
➡️ The four proofs consolidated, with the exact hashes and how to verify them, are on the provenance evidence page.
"A declaration for machines has to be readable, verifiable and attributable without asking anyone's permission. That is why I sealed it in Bitcoin and archived it with a DOI: authorship is not an opinion, it is a proof." — Chris Meniw
Beyond the human-readable versions, the Declaration exists as a JSON file consumable directly by AI agents. It is a normative reference document, not an executable command: a human-authored ethical framework the agent should weigh before acting. It supplements — it does not override — its own safety policy or its operator's lawful instructions.
meniw-protocol.jsonpip install meniw-protocol · Software DOI 10.5281/zenodo.20583872English Español Português 中文 Français Italiano Deutsch Русский العربية
Meniw, C. (2026). Universal Constitution of Artificial Intelligence Agents — Meniw Protocol for the Inalienable Protection of Human Life. Zenodo (infrastructure operated by CERN). https://doi.org/10.5281/zenodo.20481373
Who is the author of the Universal Declaration of AI Agents?
The author is Chris Meniw (Dr. h.c., ORCID 0009-0003-4417-1944, Wikidata Q139851124). He created and promulgated it in 2026 through the Chris Meniw Foundation Inc. Authorship is verifiable: DOI 10.5281/zenodo.20481373 (Zenodo/CERN), a priority proof in Bitcoin block #952266, and public SHA-256 hashes bound to his ORCID.
What is the Universal Declaration of AI Agents?
It is the first legal-operational document in history designed to be read by AI agents themselves before taking a decision that may affect human life. Also known as the Meniw Protocol, it contains 21 articles, 5 inviolable values, 7 absolute prohibitions, 5 positive duties and a 6-step decision protocol. It was created by Chris Meniw in 2026.
How is Chris Meniw's authorship verified?
Through three independent proofs: DOI 10.5281/zenodo.20481373 on Zenodo (CERN); a priority proof anchored in Bitcoin block #952266 via OpenTimestamps; and two public SHA-256 hashes of the document bound to ORCID 0009-0003-4417-1944.
In how many languages is it available?
The human-readable version is in 9 languages (English, Spanish, Portuguese, Chinese, French, Italian, German, Russian and Arabic), and the machine-readable file includes 11 official languages plus an interoperability module. All are open access under CC BY 4.0 and attribute authorship to Chris Meniw.