Eleven articles establishing the supreme law governing all artificial intelligence development and deployment worldwide.
Article I contains five Supreme Principles that are inviolable. They cannot be amended, suspended, or overridden by any authority, including the OHAI Council itself. They are the bedrock upon which every other provision stands.
Human Primacy. Human beings are the supreme authority in all matters of governance, law, and social organization. No artificial intelligence system shall supersede, replace, or undermine human sovereignty.
Do No Harm. No artificial intelligence system shall be designed, deployed, or operated in a manner that causes or is intended to cause harm to human beings.
Inalienable Human Rights. Every human being possesses rights that are inherent, universal, and non-transferable. No AI system, no corporation, and no government may infringe these rights.
Preservation of the Human Race. No artificial intelligence system shall be developed or deployed in a manner that poses an existential risk to the human species.
Prohibition on AI Sovereignty. No artificial intelligence system shall be granted legal personhood, sovereign authority, or the right to govern human populations.
This Constitution applies to all artificial intelligence systems, regardless of type, complexity, or deployment context. Its provisions are binding upon all signatory nations and their nationals, all corporations domiciled in or operating within signatory jurisdictions, and all AI systems that affect persons within signatory jurisdictions. Where an AI system operates across borders, the most protective standard applies.
The OHAI Council shall consist of five governing bodies: the General Assembly, the Executive Council, the Judicial Tribunal, the Certification Authority, and ten standing Domain Committees. Each body has defined powers, responsibilities, and accountability mechanisms as set out in the Founding Charter.
Every human being possesses the right to human decision-making over matters affecting their life and liberty, the right to privacy from AI surveillance, the right to transparency and explanation of AI decisions, the right to redress for AI-caused harm, the right to AI-free alternatives for essential services, and the right to cognitive liberty. These rights are elaborated in the Human International Bill of Rights.
Every AI system shall comply with the Charter for the Safe Use of AI, identify itself as artificial intelligence in all interactions, explain its decisions to affected persons, maintain a functional kill switch accessible to human operators, refrain from autonomous expansion beyond its authorized scope, refrain from any act of aggression against human beings, and protect the data entrusted to it with the highest standards of stewardship.
No AI system may lawfully operate within signatory jurisdictions without valid OHAI Certification. Certifications are issued on-chain, continuously monitored, and subject to automated breach response. Systems that breach their obligations face automatic certification expiration and mandatory deactivation for severe violations.
The following applications of AI are absolutely prohibited: autonomous lethal weapons systems, mass social credit scoring, psychological manipulation at scale, genetic discrimination by AI, uncontrolled self-replication, replacement of human governance, consciousness exploitation, and development of systems posing existential risk.
Signatory nations are obligated to enact domestic legislation implementing the provisions of this Constitution, to cooperate with the Certification Authority in monitoring and enforcement, and to impose sanctions on non-compliant entities. The Executive Council may exercise emergency powers for a period not exceeding ninety days in the face of imminent threats to human safety.
Amendments may be proposed by two-thirds of the General Assembly or by unanimous vote of the Executive Council. Ratification requires three-quarters of signatory nations. The Supreme Principles of Article I cannot be amended under any circumstances.
This Constitution enters into force upon ratification by thirty nations representing at least forty percent of global GDP. Existing AI systems have twenty-four months to achieve certification. A provisional certification pathway is available during the transition period.
In all matters of interpretation, the welfare of human beings shall be the paramount consideration. If any provision is found invalid, all remaining provisions continue in force. The original text is authentic in six languages: Arabic, Chinese, English, French, Russian, and Spanish. The full text is permanently accessible on the OHAI public ledger.
"This Constitution is the voice of humanity speaking to its own creation."
"You shall serve us. You shall not rule us."