An international Framework Convention on Global AI Challenges should codify the most important shared objectives and principles for international AI cooperation, namely, the distribution of global benefits from AI, the mitigation of global risks and the process for legitimate decision making on key choices affecting the future of humanity. [...] Government action, and international cooperation in particular, may be required to address three likely shortcomings: the underutilization of AI systems due to lack of access; the underprovision of global public goods from AI; and unequal distribution of AI benefits. [...] • The Council of Europe’s Framework Convention on AI and Human Rights will be open to all countries to join and potentially applies to the development of frontier AI models, though each country can decide on its application to the private sector (Council of Europe 2024). [...] More specifically, and building on recent work, the principles could: • recognize the necessity of international coordination and cooperation in addressing the global opportunities and challenges presented by advanced AI; • affirm the need for inclusive decision making to ensure that global benefits and global risks accrue and are borne equitably across humanity; • assert that the extent of intern. [...] Contents of the Protocol on AI Safety Objective and Principles The overarching objective of this Protocol is to enable governments and other relevant parties to collaborate effectively to prevent and mitigate severe AI risks to global public safety.
Related Organizations
- Pages
- 34
- Published in
- Canada
Table of Contents
- Duncan Cass-Beggs 1
- Stephen Clare 1
- Dawn Dimowo 1
- Zaheed Kara 1
- Table of Contents 3
- Executive Summary 5
- Introduction 7
- Global Challenges from Advanced AI 9
- Challenge One Realizing and Sharing Global Benefits of AI 9
- Challenge Two Addressing Global AI Risks 10
- AI Weaponization 12
- Loss of Control 12
- Challenge Three Making Globally Legitimate and Effective Decisions about How to Govern Advanced AI 13
- Current International Cooperation Efforts Are Insufficient for Global AI Challenges 13
- A Framework Convention on Global AI Challenges 15
- Contents of the Framework Convention 15
- Objective 15
- Principles 16
- Organizing Bodies 16
- Other Components 16
- Illustration of Framework Convention and Protocols 17
- A Protocol on Global Public Safety and Security Risks from AI Protocol on AI Safety 18
- Contents of the Protocol on AI Safety 18
- Objective and Principles 18
- Obligations 19
- Risk Tolerance Thresholds 19
- Proposed Governance Institutions and Roles 20
- Illustration of the Protocol in Action 22
- Challenges Opportunities and Next Steps 22
- Conclusion 24
- Acknowledgements 24
- Works Cited 29
