The report, "Governing AI for Humanity", emphasizes that AI's rapid advancements pose significant risks, including widening inequalities, disinformation, and threats to global security, unless a comprehensive, internationally coordinated governance framework is implemented.

Global Gaps and Risks

The report identifies major gaps in current AI governance, especially in representation, coordination, and accountability. Despite a proliferation of national and regional AI frameworks, many countries, particularly in the Global South, are excluded from governance discussions. This exclusion risks leaving entire regions behind, as AI-related benefits remain concentrated in few nations and powerful corporations.

Additionally, the report outlines growing risks associated with AI, including biases in AI systems, the spread of disinformation through AI-generated content, and the potential for AI to exacerbate inequality by concentrating power and wealth. The report also raises concerns about AI’s role in autonomous weapons systems, posing a direct challenge to global peace and security.

To address these challenges, the advisory body calls for a globally coordinated approach to AI governance giving key recommendations.

Creation of an International Scientific Panel on AI

This body would consist of diverse AI experts, supported by the UN AI office, tasked with publishing an annual report on AI trends, risks, and research gaps, producing quarterly updates on AI’s role in achieving the SDGs, and issuing reports on emerging risks and governance gaps.

Launch of a Policy Dialogue on AI Governance

Holding a twice-yearly intergovernmental and multi-stakeholder policy dialogue on AI governance, aimed at sharing best practices that promote development and human rights, enhancing international interoperability of AI governance, discussing significant AI incidents, and reviewing reports from the international scientific panel on AI.

AI Standards Exchange

The creation of an AI standards exchange, bringing together representatives from standard-development organizations, tech companies, civil society, and the international scientific panel to develop and maintain a register of AI standards, evaluate existing standards and processes, and identify gaps where new standards are needed.

Capacity Development Network

The creation of an AI capacity development network to link up a set of collaborating, United Nations-affliated capacity development centres making available expertise, compute and AI training data to key actors. Its goals would include aligning global AI capacity efforts, building AI governance capacity for public officials, and offering resources like trainers, computing power, and educational opportunities to researchers and social entrepreneurs. The network would also support cross-disciplinary teams, create sandboxes for testing AI solutions, and offer a fellowship program for promising individuals.

Global fund for AI

This fund would create a global fund for AI to bridge the AI divide, managed by an independent governance structure. It would receive contributions from public and private sources to facilitate access to AI resources through the capacity development network. The fund would support shared computing resources, sandboxes for testing AI solutions, governance and safety tools, data sets for SDG-related projects, and a repository of AI models and curated data sets.

AzVision.az

More about: