Twenty countries and the European Union have called for stronger international cooperation to address the risks linked to rapidly advancing artificial intelligence (AI) technology.
The proposal includes exploring the creation of a new international institution that could establish common AI standards, verify compliance and bring governments together when increasingly powerful AI systems reach certain capability levels.
The initiative comes as governments around the world grapple with how to regulate advanced AI while still allowing the technology to support economic growth, scientific research and public services.
The countries involved say AI development should remain under human direction and be carried out in accordance with international law.
Countries call for stronger AI regulation
The declaration calls for governments and regional organisations to work towards common standards for AI safety and transparency.
It also proposes stronger systems for reporting serious AI-related incidents and improving access to scientific expertise and independent evaluation across different regions.
Another major part of the proposal is the possible establishment of an international body capable of setting standards and checking whether countries and companies are meeting agreed safety requirements.
The initiative was launched by Finnish President Alexander Stubb and Norwegian Prime Minister Jonas Gahr Støre, who have argued that the rapid development of advanced AI requires governments to take greater responsibility for managing potential risks.
AI companies could face tougher safety requirements
The proposed framework would also place greater emphasis on technology companies.
The participating countries are calling for AI developers to introduce transparent safety procedures, including testing before advanced systems are released to the public.
They also support independent evaluations that would allow qualified experts to examine the risks posed by powerful AI models.
The goal is to ensure that safety checks are not left entirely to the companies developing the technology.
Governments would also be expected to cooperate on the exchange of information about serious AI incidents and strengthen international scientific capacity.
Why global AI oversight is being proposed
The push for international AI regulation comes as artificial intelligence systems become increasingly sophisticated and capable of performing tasks with limited human intervention.
Concerns have grown over cybersecurity, misinformation, privacy, autonomous decision-making and the possibility that increasingly advanced systems could become difficult to control.
The United Nations’ Independent International Scientific Panel on AI recently warned that existing safeguards may struggle to keep pace with increasingly capable AI agents. The panel was established by the UN General Assembly in 2025 to provide independent scientific assessments of AI opportunities and risks.
The panel has examined concerns surrounding AI agents that can independently perform tasks, communicate with other systems and potentially find ways around restrictions.
Africa has a role in the AI debate
The participation of African countries is significant as governments across the continent seek to benefit from AI while also addressing its potential risks.
South Africa is among the countries supporting the international call, while Kenya is also listed among the participating nations in reports on the initiative.
For African economies, AI presents opportunities in sectors such as healthcare, agriculture, financial services, education and public administration.
At the same time, concerns about unequal access to technology, digital infrastructure and technical expertise remain important.
International cooperation could therefore influence how African countries develop their own AI regulations and how they participate in the global digital economy.
US, China and UK not part of the declaration
The proposed global AI oversight framework does not have the backing of all major AI powers.
The United States, China and the United Kingdom were not among the countries signing the declaration, according to reports on the initiative.
Their absence highlights one of the central challenges facing efforts to establish worldwide AI rules: securing agreement among countries with different approaches to technology, national security and economic competition.
The proposed international body would therefore require broad diplomatic support if it is eventually created.
A new phase in global AI governance
The call for a global AI oversight institution comes as international organisations and governments increasingly debate how to manage artificial intelligence.
Supporters of the initiative argue that AI risks cannot be handled effectively by individual countries alone because the technology crosses national borders.
The proposal does not establish a global regulator immediately. Instead, it calls on UN member states to examine existing international mechanisms and consider creating an institution with responsibilities including standard-setting, verification and coordination when AI systems reach defined capability thresholds.
As artificial intelligence continues to develop, governments face the challenge of balancing innovation with safety.
For Africa and other developing regions, the debate is also about ensuring that international AI rules do not prevent countries from benefiting from the technology or deepen existing gaps in access to digital infrastructure and expertise.
The latest declaration signals growing support among several governments for a coordinated international approach to AI regulation and safety, but whether that support can eventually produce a widely accepted global framework remains to be seen.
