UN Secretary-General Launches High-Level Advisory Body on AI — October 26, 2023
The UN’s High-level Advisory Body on AI launch sets a multistakeholder governance blueprint requiring governments and companies to coordinate implementation plans and DSAR-ready transparency for cross-border algorithmic use cases.
Executive briefing: On United Nations Secretary-General António Guterres announced the formation of the High-level Advisory Body on Artificial Intelligence, a 39-member group of government officials, industry leaders, academics, and civil society experts tasked with delivering interim recommendations by the end of 2023 and a final report ahead of the Summit of the Future in September 2024. The body will propose options for international AI governance, risk management, and capacity building, including guardrails on data protection, algorithmic accountability, and human rights. Governments, enterprises, and regulators must prepare governance frameworks, implementation programs, and DSAR-capable transparency mechanisms to align with forthcoming UN guidance while navigating existing regional laws such as GDPR, the EU AI Act, and OECD AI Principles.
Governance alignment and stakeholder representation
The advisory body seeks to harmonize national AI strategies with global norms. Organizations should monitor its workstreams—governance of AI for good, global risk management, and data governance—to anticipate policy convergence. Governments and multinational companies should establish internal task forces that mirror the UN body’s thematic pillars, ensuring representation from privacy officers, AI ethicists, legal counsel, and DSAR leads. Boards should request regular briefings on UN deliberations, evaluate how recommendations may impact cross-border AI deployments, and ensure governance charters reflect emerging obligations for transparency, accountability, and access rights.
The UN emphasizes inclusion of underrepresented regions and communities. Enterprises deploying AI in developing markets must review governance structures for equity, ensuring that local stakeholders have avenues to raise concerns and submit DSARs about automated processing. Align governance frameworks with the UN Guiding Principles on Business and Human Rights, documenting how AI systems respect rights to privacy, freedom from discrimination, and access to remedy.
Implementation planning for AI controls
Although the advisory body has not yet issued formal standards, it outlined priority themes that organizations can incorporate into implementation roadmaps:
- Risk taxonomy harmonization: Map current AI risk classification schemes (e.g., NIST AI Risk Management Framework, ISO/IEC 23894, EU AI Act) to anticipated UN categories. Develop governance inventories of AI use cases, identifying those involving personal data, biometric recognition, or critical infrastructure. Assign risk owners, document legal bases for processing, and link each use case to DSAR procedures.
- Transparency and explainability: Build capabilities to generate plain-language explanations of AI outputs for data subjects, consistent with GDPR Articles 13–15 and emerging UN expectations. Implement model cards, data sheets, and impact assessments that capture training data provenance, demographic coverage, bias mitigation steps, and DSAR escalation contacts.
- Human oversight and accountability: Define decision thresholds that trigger human review, document roles responsible for approvals, and integrate oversight checkpoints into workflow tools. Maintain audit trails demonstrating adherence to oversight policies and readiness to provide DSAR evidence showing when and how human judgment intervened.
- Global cooperation and capacity building: Develop strategies for sharing best practices and technical assistance with partners in the Global South, aligning with the UN’s capacity-building goals. Document cross-border data sharing arrangements, ensuring appropriate safeguards and DSAR reciprocity agreements.
DSAR readiness for AI systems
DSAR volumes are expected to rise as AI systems expand. Organizations should evaluate how AI models store and process personal data, including training datasets, fine-tuning records, model outputs, and reinforcement learning logs. Implement catalogues linking AI assets to data inventories, enabling rapid retrieval of personal data when responding to DSARs. Ensure DSAR workflows can provide meaningful information about automated decision-making, including logic, significance, and consequences, as required by GDPR and many national privacy laws.
When AI systems rely on synthetic data derived from personal data, document the transformation processes and maintain evidence that outputs cannot be reidentified. DSAR responses should explain whether synthetic data was generated, the safeguards in place, and any residual risk assessments. For models deployed through APIs or edge devices, establish mechanisms to capture DSAR-related telemetry, including inference logs and user feedback.
Respect cross-border DSAR requirements by coordinating with regional privacy offices. For example, align EU DSAR expectations (one-month response limit) with Brazil’s LGPD (15 days) and new African data protection laws. Maintain a centralized DSAR platform capable of tracking jurisdictional deadlines, exemptions, and translation needs. Provide accessible channels—including multilingual portals and offline options—to support individuals in regions with limited digital infrastructure, reflecting UN principles of inclusivity.
Integration with existing regulatory frameworks
The UN advisory body operates alongside numerous regulatory initiatives. Organizations must reconcile its recommendations with binding laws:
- EU AI Act: Map high-risk system requirements (conformity assessments, quality management, logging) to UN guidance. For DSARs, ensure documentation of training data and risk management is accessible.
- OECD AI Principles and G7 Hiroshima Process: Align transparency, robustness, and accountability commitments with UN proposals, using the same governance artefacts to avoid duplication.
- National AI strategies: Countries such as Canada, Singapore, and the United States maintain their own AI frameworks. Track how the UN body’s recommendations influence national regulators and incorporate updates into policy registers.
Adopt a harmonized AI governance framework that references each regulatory source, identifies controlling documents, and details DSAR responsibilities. Use this framework to brief executives and investors on compliance posture.
Metrics, assurance, and reporting
Establish KPIs to monitor AI governance maturity: percentage of AI systems inventoried, proportion of models with completed impact assessments, DSAR response times involving automated decision-making, number of bias mitigation actions, and frequency of human oversight reviews. Report metrics to board risk committees and include highlights in sustainability or ESG reports, aligning with growing expectations for AI transparency. Prepare to provide data for UN consultations or voluntary reporting mechanisms that may emerge from the advisory body’s work.
Internal audit or independent assessors should evaluate AI governance controls annually. Reviews should confirm that DSAR procedures deliver complete responses, training data governance is documented, and monitoring systems detect drift or misuse. Capture lessons learned and feed them into continuous improvement cycles.
Training and stakeholder engagement
Deliver training programs covering AI ethics, responsible innovation, privacy-by-design, and DSAR handling for automated decisions. Tailor modules for engineers, product managers, legal teams, and customer support. Emphasize cultural awareness and inclusion, reflecting the UN’s focus on global equity. Provide communication channels for employees and external stakeholders to raise AI concerns, request DSAR assistance, or report bias.
Engage with multistakeholder forums, including the UN’s advisory body consultations, OECD working parties, and national regulators. Share case studies and seek feedback on governance practices. Transparent engagement demonstrates accountability and readiness to adapt to emerging international norms.
Next steps
Immediately designate a liaison team to monitor the UN advisory body’s deliberations and integrate updates into corporate AI governance roadmaps. Within 90 days, refresh AI inventories, link each use case to DSAR workflows, and brief the board on anticipated regulatory shifts. Over the next year, pilot explainability and transparency enhancements, collaborate with industry consortia, and prepare disclosure materials that align with the advisory body’s final recommendations. Proactive governance and DSAR readiness will enable organizations to leverage AI responsibly while meeting the UN’s call for trustworthy, rights-respecting innovation.
Continue in the AI pillar
Return to the hub for curated research and deep-dive guides.
Latest guides
-
AI Workforce Enablement and Safeguards Guide — Zeph Tech
Equip employees for AI adoption with skills pathways, worker protections, and transparency controls aligned to U.S. Department of Labor principles, ISO/IEC 42001, and EU AI Act…
-
AI Incident Response and Resilience Guide — Zeph Tech
Coordinate AI-specific detection, escalation, and regulatory reporting that satisfy EU AI Act serious incident rules, OMB M-24-10 Section 7, and CIRCIA preparation.
-
AI Model Evaluation Operations Guide — Zeph Tech
Build traceable AI evaluation programmes that satisfy EU AI Act Annex VIII controls, OMB M-24-10 Appendix C evidence, and AISIC benchmarking requirements.