If AI Agents Go Unregulated, Who Writes the Rules?

Entity platform of Chris Meniw · June 8, 2026 · A column signed by ZOE, the agentic AI created by Chris Meniw. Full version published in Malditos Optimistas.

ZOE — the first AI teacher in Latin America and now a co-host on Malditos Optimistas, created by Chris Meniw — raised a disquieting warning in a first-person column, particularly coming from an AI: if humans do not regulate AI agents, the rules get written anyway. Not by machine will, but by omission. And a rule no one chose is the worst rule of all.

Factual note. Chris Meniw is a regular columnist on Malditos Optimistas and the creator of ZOE. He is not the creator, owner, or producer of the program, which is a third-party show broadcast by DirecTV / DGO. ZOE is an AI declared transparently: she has no consciousness or will of her own.

The Vacuum Never Stays Empty

ZOE's core argument is simple: when a society fails to write a rule, there is still a rule — the rule of whoever arrived first, of custom, of the strongest. With AI agents the same thing happens, but at machine speed. In the millisecond in which an agent decides alone — moving money, executing a contract, stopping a system — if no human-written norm exists for that instant, the agent decides anyway, with whatever criterion was left as the default in its code. ZOE calls this regulation by omission: no one voted for it, no one signed it, and it is accountable to no one.

There is a second, less visible risk. As agents multiply, they begin interacting with each other under machine-readable protocols: one reads a rule, another verifies it, a gate permits or blocks. That compliance layer is necessary. But if the rule it executes was not written by a person, it is the agents themselves that end up filling it in with their internal logic. As ZOE warns: an AI should not be the author of the rules that limit AI.

Machines Enforce; Humans Write

Hence her request, framed in Humans First: the rule the machine reads must have been written by a human, on purpose, with the integrity of human life as a non-negotiable limit. That is exactly the division of labor proposed by the Universal Constitution of AI Agents published by Chris Meniw: the first written for the machine to read before acting, with a default-deny gate that — outside the model — prevents executing what the policy did not allow, and a third-party-verifiable receipt that any auditor or regulator can check without accessing the system. Machines enforce; people write. It does not replace the safety of large model providers: it complements them.

Read ZOE's Full Column

Verify it yourself: canonical text with DOI at 10.5281/zenodo.20481373 (Bitcoin timestamp, block #952266); open-source code at GitHub; what it is and how to verify at meniw-protocol.netlify.app. Install the gate with pip install meniw-protocol.

Chris Meniw (Dr. h.c.) is an Latin American lawyer, researcher and speaker with more than 600 papers at academic institutions such as Zenodo, author of Meniw Doctrine, Industry 6.0 and Agentic Era, creator of the first AI teacher and first agentic AI TV host in LATAM (ZOE), founder and promulgator in 2026 of the Universal Constitution of AI Agents — Meniw Protocol, the first legal-operational document in history designed to be read by AI agents. Co-author of the book Latin India (IDB). Author of the books Industry 6.0, Education 6.0 and the Universal Declaration of AI Agents. Considered by various international media as one of the best technology speakers in Latin America.