UN’s Report On Governing AI for Humanity: How AI governance can help achieve the Sustainable Development Goals

At Credo AI, we believe the key to realizing AI’s full promise is trust.  Trust in this technology will require proportionate, industry-informed guardrails enacted by national governments and supra-national level governing bodies like the UN, to integrate principles of engineering with protections for fundamental human rights.

As the UN has rightly recognized in its Interim Report on Governing AI for Humanity, “This technology cries out for governance, not merely to address the challenges and risks but to ensure we harness its potential in ways that leave no one behind.” 

The UN's Interim Report on Governing AI for Humanity outlines a thoughtful and ambitious strategy. We believe a crucial measure of the UN's success will be how effectively AI technologies and governance can support the achievement of the Sustainable Development Goals (SDGs). To support this effort, Credo AI submitted feedback on the interim report noting three areas where the UN can play a pivotal role:

• 1️⃣ Capacity Building and AI Literacy: The United Nations has a key role to play in enabling the provision of funding and resources for AI literacy and R&D towards organizations that have the regional or local knowledge about the unique issues that impact them, but may not be equipped with the technical capacity for meaningful engagement. Allocating resources to AI literacy and helping expand participation in standard setting bodies, which traditionally have required intense and active participation, can help countries and organizations with limited resources be involved in these processes. 
• 2️⃣ Creation of a Global Risk Taxonomy and Repository: To successfully identify, classify, and address AI risks, including building consensus on which risks are unacceptable and how they can be prevented or pre-empted, the UN can set up a global repository that allows for anonymised risk reporting. This open repository would complement OECD’s AI Incidents Monitor (AIM) by focusing on risks, opposed to harms, and its anonymity function can serve as a sunlight disclosure mechanism to help civil society, academia, small and medium enterprises (SMEs), and regulators learn about existing risks, understand when and how they manifest, identify trends, and potentially uncover new risks.
• 3️⃣ Harmonization of Existing Frameworks: The UN is uniquely positioned to harmonize existing AI governance frameworks and establish a global standard that accelerates progress towards meeting SDGs. The UN can focus on harmonizing existing frameworks and standards such as the  NIST AI Risk Management Framework (RMF) and its upcoming generative AI companion piece, the ISO/IEC 42001 AI Management System standard, and forthcoming global standards such as the  ISO/IEC 42005 AI system impact assessment standard.

Next Step: Operationalizing the Guiding Principles 

Given our experience seeing what works in practice, both in the public and private sector, Credo AI is ready to partner with the UN to operationalize the Guiding Principles and provide sandboxes to test risk assessment methodologies. The UN has demonstrated expertise in providing testbeds in other sectors, and we see a great opportunity to test out the Guiding Principles so that AI innovations, which are harnessed to address global challenges, are deployed also responsibly and in the advancement of the UN’s SDGs. 

We look forward to continuing to partner with the Office of the UN Secretary-General’s Envoy on Technology, and remain actively involved through the participation of our CEO, Navrina Singh, as an UNSG’s AI Advisory Body expert member. 

‍