Blog
AI Policy, Global Governance, Regulation

22 Countries AI Oversight Proposal: What the International Declaration Says

September 23, 2026
time
22 Countries AI Oversight Proposal: What the International Declaration Says
WRITTEN BY
GlobalNodes
IN THIS ARTICLE

A coalition of 22 countries has backed a declaration calling for stronger international oversight of artificial intelligence.

The initiative was discussed during the 2026 UN General Assembly and calls for AI systems to remain under meaningful human control. The United States and China did not sign the declaration.

What are the countries proposing?

The declaration focuses on creating stronger international safeguards around advanced AI.

Among the proposals are calls for AI companies to adopt transparent safety protocols, conduct pre-deployment testing and allow independent evaluations.

The broader goal is to establish common expectations for AI safety across countries.

Why is international oversight being discussed?

AI companies operate globally.

A model trained in one country can be deployed in another within minutes.

AI agents can also interact with international infrastructure, cloud platforms and software services.

That creates challenges for purely national regulation.

One country's rules may not address risks created by an AI system operating across multiple jurisdictions.

What could international oversight look like?

A global framework could potentially involve:

Common AI safety standards

Independent testing

Incident reporting

Cross-border cooperation

Model transparency

Human oversight requirements

However, the declaration does not itself create a global regulator.

Creating an international institution would require further negotiations between governments.

What does this mean for businesses?

Global AI companies may eventually have to operate under multiple layers of governance.

That makes flexible compliance infrastructure increasingly important.

Organizations should be able to document how models are trained or sourced, what data they use, what permissions they receive and how incidents are handled.

The 22-country proposal is therefore another sign that AI governance is moving from individual national policies toward a broader international discussion.

Whether that results in a formal global oversight body remains to be seen.

Ready to start your project?

Have a project in mind? We'd love to hear about it. Tell us what you're building and let's explore what's possible.

Email

hello@globalnodes.com

WhatsApp

+91 9873388887

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.