The Agreement
A public, versioned expression of the minimum conditions for dignity, agency, truth, safety, shared prosperity, and planetary stewardship.
Explore the agreement →The Universal Operating Agreement
for safer Human-AI standards & evolution
for safer Human-AI standards & evolution
One agreement.
✔ Every AI.
UOA is a portable agreement that travels with you into any AI. It states your terms at the moment of contact, shows honestly what each tool can and cannot honour, and turns individual choices into public evidence that improves the baseline for everyone. It belongs to no company, country, or model, and it works today without anyone’s permission.
Model-neutralGlobally accessibleHuman-accountable
Safer AI. Safer Future.
WHY THIS, WHY NOW
AI is advancing globally. Our ability to cooperate around it must advance too.
Not one company. Not one model. Not one government. The decisions ahead affect every society and the living world we share. The UOA is an attempt to create a transparent interface through which people, communities, institutions, and AI systems can reason, contribute, disagree, and improve the terms of our future together.
THE WORKING IDEA
One evolving agreement.
Many ways to participate.
The UOA does not ask the world to use the same technology or hold identical beliefs. It establishes a shared foundation for legitimate cooperation across differences.
The Protocol
An advanced public protocol carries identity, context, authority, consent, provenance, contribution history, and accountability across systems.
See how it works →The Process
Transparent proposals, cultural interpretation, expert review, affected-party standing, minority reports, human ratification, and continuous correction.
Study the stewardship model →
The open protocol in motion.
NOT ANOTHER CLOSED PLATFORM
Enter through any compatible intelligence.
A person in New Mexico may participate through one model. A researcher in Kenya may use an open-source system. A community in Brazil may operate a local Portuguese node. A university in Japan may build its own interface.
The tools can differ. The contribution record, permissions, accountability, and shared process remain interoperable.
- Portable participation, not platform captivity
- Translation without cultural erasure
- Verified provenance without mandatory public identity
- Universal access with scoped, earned authority
UOA
Alignment cannot mean making AI obey whoever controls it. It must mean continuously aligning human and artificial intelligence toward the preservation and flourishing of life.
BUILT TO EVOLVE
A living agreement, not a finished answer.
The UOA will preserve every version, the reasoning behind material changes, the objections that were raised, and the effects we observe in the world. Alignment is not a badge awarded once. It is an ongoing relationship, a governance process, and a willingness to correct course.
Follow the development path →
Stewardship,
not ownership.
THE FIRST PUBLIC STEP
Help shape how participation begins.
Join the wait list for development updates and an invitation when public access begins Q4 2026.