Norwegian Prime Minister Jonas Gahr Støre and 22 other world leaders have signed a political declaration aimed at promoting safe artificial intelligence development. However, legal and technological experts have criticized the pact for lacking real enforcement power, binding legal commitments, and sanction mechanisms. The initiative pushes for transparent safety protocols and discusses a potential UN institution for AI, yet major industry players in the United States and China remain outside the framework.
Global Leaders Sign Safety Declaration Without Enforcement Powers
Prime Minister Jonas Gahr Støre joined 22 international leaders to sign the political appeal for safe artificial intelligence development. Notable signatories include European Commission President Ursula von der Leyen, German Chancellor Friedrich Merz, Danish Prime Minister Mette Frederiksen, Finnish President Alexander Stubb, and Canadian Minister Mark Carney. According to the Prime Minister’s Office in Norway, the coalition spans multiple continents, incorporating leaders from Australia, Singapore, South Africa, and the United Arab Emirates.
The political declaration outlines three primary objectives for participating governments and industry actors. First, it urges companies developing AI systems to establish transparent safety protocols, including mandatory testing and independent evaluations to measure risk. Second, it calls for national authorities and regional bodies to coordinate common standards and build a system for sharing severe safety incidents across borders. Third, the pact aims to ensure that participating countries gain access to research capacity, technical expertise, and credible evaluations.
The initiative also proposes that United Nations member states explore the creation of an international AI institution. This proposed body would build on existing international frameworks to set technical standards, enable verification, and address states or developers that cross established safety boundaries.
Legal and Technical Critics Point to Regulatory Gaps
Despite the high-profile list of signatories, legal and technical experts argue that the declaration lacks practical authority. Malcolm Langford, a professor of law at the University of Oslo, pointed out that the document contains no binding legal requirements. According to Langford, the initiative functions purely as a political promise to draft rules in the future rather than a treaty carrying legal obligations or penalties.
Morten Goodwin, a professor of artificial intelligence at the University of Agder and a member of the Norwegian Board of Technology, echoed these concerns regarding the pact’s lack of muscle. Goodwin noted that a signature from a national leader does not dictate the research or development priorities of private technology firms. He suggested that European politicians are focusing heavily on regulatory frameworks rather than driving domestic innovation, while the competitive race for advanced AI capabilities accelerates globally.
The world's leading AI nations—the United States and China—have not signed the declaration. Major American technology companies, such as Google and investors behind Anthropic, operate with market valuations scaling into the trillions of dollars, far outpacing Europe’s largest artificial intelligence firm, the French company Mistral, which holds a market valuation of roughly $24 billion.
Next Steps for International AI Governance
Prime Minister Støre has acknowledged that major AI developer nations must be brought into future cooperative efforts if international oversight is to succeed. Yet, researchers remain skeptical that voluntary declarations will alter the pace of development. Goodwin noted that despite public statements from companies like Google and investors behind Anthropic regarding the need to slow down or carefully manage their releases, major AI models continue to debut at a rapid rate, leaving governments working to catch up to an industry moving faster than traditional diplomacy.
>Worth a look