IAIO Governance Framework: Consultation Draft

There is hereby established an International Artificial Intelligence Organization (IAIO), constituted as an intergovernmental organization with international legal personality, established by multilateral treaty among member states.

The objectives of the IAIO shall be to promote international cooperation on AI governance; to establish and maintain international standards for AI safety, transparency, and accountability; to facilitate the sharing of information about AI risks, incidents, and best practices; to support capacity-building in AI governance among member states; and to ensure that the development and deployment of AI serves the common good of humanity.

The mandate of the IAIO shall extend to artificial intelligence systems, autonomous systems, robotics, and frontier intelligence technologies as defined by the Technical Secretariat in accordance with guidelines adopted by the General Assembly.

Membership of the IAIO shall be open to all states that are members of the United Nations. States may accede to the founding treaty at any time following its entry into force.

Observer status may be granted by the General Assembly to non-member states, intergovernmental organizations, and civil society organizations meeting criteria established by the General Assembly.

The General Assembly, comprising all member states, shall be the supreme governing body of the IAIO. It shall meet in regular session annually and in special session as required. Each member state shall have one vote. Decisions on matters of substance shall require a two-thirds majority; procedural matters shall be decided by simple majority.

The Executive Board shall comprise 35 member states elected by the General Assembly for three-year terms, with equitable geographic representation. The Board shall be responsible for the implementation of General Assembly decisions and the oversight of the Technical Secretariat.

The Technical Secretariat, headed by a Director-General appointed by the General Assembly, shall provide the expert and administrative capacity of the IAIO. The Secretariat shall be staffed by international civil servants of the highest standards of competence, integrity, and independence.

The IAIO shall develop, adopt, and maintain international standards for AI safety, transparency, accountability, and responsible development. Standards shall be developed through an inclusive, evidence-based process involving member states, technical experts, civil society, and affected communities.

The IAIO shall establish and maintain mechanisms for the sharing of information among member states about AI risks, incidents, vulnerabilities, and best practices, subject to appropriate safeguards for sensitive information.

The IAIO shall provide technical assistance and capacity-building support to member states, with particular attention to developing countries, to enable effective participation in AI governance and responsible AI development.

The IAIO shall develop and implement mechanisms for the oversight of AI development and deployment by member states and, where authorized, by non-state actors. Oversight mechanisms shall be proportionate, evidence-based, and respectful of national sovereignty.

The IAIO shall be financed by assessed contributions from member states, calculated according to a scale adopted by the General Assembly, and by voluntary contributions from member states and other donors. The budget shall be adopted by the General Assembly by a two-thirds majority.

The IAIO shall be guided in all its activities by the principles of: human dignity and human rights; international peace and security; sustainable development; equity and inclusiveness; transparency and accountability; scientific integrity; and the rule of law.