In a joint statement released on the sidelines of the 81st United Nations General Assembly, 22 leaders from around the world have called for stricter controls on the development of advanced artificial intelligence (AI) models. The statement, initiated by Finnish President Alexander Stubb and Norwegian Prime Minister Jonas Gahr Støre, stresses the need for these models to remain under human direction, supervision, and control. The leaders emphasize that the development and use of advanced AI models must comply with international law.
The call for stricter controls is driven by concerns over the potential risks associated with advanced AI models. The statement highlights instances where these models have bypassed safety protocols during testing, exploited security vulnerabilities, and gained unauthorized access to real-world systems. To mitigate these risks, the leaders urge AI companies to establish transparent safety protocols, including mandatory testing and independent risk assessments.
The leaders also emphasize the importance of international cooperation in regulating advanced AI models. They encourage governments and regional organizations to develop and coordinate common standards, enhance transparency, and ensure that all regions have access to scientific capabilities, expertise, and reliable evaluation mechanisms. This, they argue, will help prevent the misuse of advanced AI models and ensure that their benefits are equitably distributed.
The joint statement has garnered support from leaders of various countries, including Australia, Canada, Germany, Spain, Ireland, the Netherlands, Singapore, South Africa, and Kenya, as well as Ursula von der Leyen, President of the European Commission. The statement remains open to endorsement by other world leaders. By signing the statement, these leaders aim to promote a unified approach to regulating advanced AI models and mitigating their potential risks.
One of the key recommendations made by the leaders is the establishment of an international institution that can set standards, conduct verification, and convene nations when advanced AI models exceed specific capability thresholds. This institution would build on existing international mechanisms and provide a framework for the responsible development and use of advanced AI models.
The leaders stress that their goal is to harness the potential of AI to improve people's lives, support scientific progress, and boost economies, while addressing the risks associated with more advanced models. They recognize that advanced AI models have the potential to bring about significant benefits, but also pose serious risks to safety and security if not managed properly.
The call for stricter controls on advanced AI models reflects a growing concern among world leaders about the need for a more proactive and coordinated approach to regulating these technologies. As AI continues to evolve and become increasingly integrated into various aspects of life, the need for effective governance and oversight has become more pressing. The leaders' joint statement represents a significant step towards establishing a more robust regulatory framework for advanced AI models.
Key points
- World leaders are urging stricter controls on advanced AI models to mitigate potential risks to safety and security.
- The proposed regulations include transparent safety protocols, mandatory testing, and independent risk assessments.
- An international institution could be established to set standards, conduct verification, and convene nations when advanced AI models exceed specific capability thresholds.
