Human Rights and International Law Should Guide AI Panel and Dialogue
Overview of the UN Consultation on AI Governance
On April 2, 2025, a significant consultation regarding the Zero Draft for the AI Scientific Panel and Global Dialogue on AI Governance took place at the United Nations. A coalition of organizations, including ARTICLE 19, and experts in international relations, human rights, and technology, has written to the Permanent Representatives of Costa Rica and Spain. They emphasize the need to address crucial shortcomings in the draft to ensure that the governance of artificial intelligence (AI) is grounded in human rights, inclusive, transparent, and capable of mitigating serious risks associated with AI technologies.
Concerns About Procedural Aspects
Government-Centric Structure
One of the major issues raised by the coalition pertains to the proposed government-centric approach in selecting the panel’s composition. Specifically, paragraphs 3(b), 4, and 5 highlight concerns regarding how this approach may lead to corporate influence or state dominance over the process. The members of the coalition stress the necessity for solid safeguards to prevent such outcomes. Additionally, they propose that the disclosure of any financial, professional, or personal interests affecting impartiality should not only be part of the initial nomination process but also an ongoing requirement for the panel.
Clarity on Panel Composition
The structure of the panel consists of two bodies: an Advisory Committee elected by the General Assembly and an Expert Committee. However, there is insufficient clarity on how these committees will interact and influence each other’s work. The lack of clear guidelines could lead to an opaque process, creating room for political maneuvering that may undermine the panel’s objectives. Experts may be reluctant to serve pro bono if they perceive the panel as politically influenced.
Defining the Scope of the Dialogue
The dialogue aims to complement existing policy discussions on AI to avoid overlaps with other ongoing efforts. The coalition argues that it should function alongside the UN Internet Governance Forum (IGF). There are concerns regarding the characterization of the dialogue as both "multistakeholder" and "intergovernmental." This duality could risk turning the dialogue into another regulatory framework rather than a platform for genuine discussion.
Geopolitical Considerations
The anticipated first dialogue is scheduled to take place in New York, which raises concerns given the geopolitical context affecting civil society, particularly in the Global Majority countries. The coalition suggests reconsidering the venue for future dialogues to ensure a more inclusive and secure environment, such as Geneva.
Emphasizing Civil Society Participation
The absence of a structured role for civil society is another significant concern highlighted by the coalition. Civil society members, who will be most affected by AI technologies, need guaranteed opportunities for meaningful participation. Without clearly defined mechanisms for their involvement, existing power imbalances are likely to persist, undermining the multistakeholder approach crucial for effective governance.
Human Rights Framework and AI Governance
The coalition insists that the mandates for both the panel and the dialogue must be grounded in international law and human rights frameworks. While human rights are briefly mentioned in the Zero Draft, the document fails to detail how these principles will be implemented throughout the lifecycle of AI technologies. For instance, high-risk applications of AI, such as predictive policing, are not adequately addressed. The coalition advocates that existing human rights frameworks, like the UN Guiding Principles on Business and Human Rights, must be incorporated into the discussions.
Protecting Scientific Integrity
For the panel to function effectively, it must balance scientific integrity with political legitimacy. Ensuring the credibility of the panel’s outputs is crucial, and this can only be achieved through a robust human rights framework. Without this, the governance initiatives on AI may lack coherence and effectiveness.
Recommendations for Improvement
The coalition urges member states to take the necessary steps to ensure that the AI Scientific Panel and Global Dialogue are well-structured, focused on human rights, and responsive to the challenges posed by AI technologies. They call for eliminating vague language and ambiguities, advocating for a clear and actionable framework that genuinely reflects the needs of all stakeholders in a rapidly changing digital landscape.
Signatories
The letter has garnered support from various organizations, including Access Now, ARTICLE 19, and the Association for Progressive Communications, along with notable individuals from academic institutions and think tanks, emphasizing the widespread concern about the future of AI governance.
