Both the UNESCO Recommendation on the Ethics of AI Implementation Program and the Partnership on AI's AI and Shared Prosperity Initiative offer valuable frameworks for promoting ethical AI development and deployment. The UNESCO Recommendation provides a broad, government-focused approach aligned with international law, while the Partnership on AI's initiative focuses on the economic impacts of AI and targets a wider range of stakeholders. The choice between them depends on the specific goals and context of the user.
| Attribute | UNESCO Recommendation on the Ethics of AI Implementation Program | Partnership on AI's AI and Shared Prosperity Initiative |
|---|---|---|
| Scope of ethical guidelines covered | Addresses human rights, fairness, accountability, transparency, privacy, and the promotion of the public good. Emphasizes key policy areas such as data governance, gender equality, and AI applications in education, culture, labor markets, the environment, communication, information, health, social well-being, and the economy. | Guidelines for ensuring AI contributes to shared prosperity and mitigates potential negative impacts on job markets; algorithmic equity; fairness, transparency, and accountability; responsible use of AI; sustainable and responsible innovation. |
| Target audience | Primarily addressed to Member States, but provides ethical guidance to all AI actors, including public entities, private sector companies, and civil society organizations. | AI developers, AI users, policymakers, labor organizations, and workers. |
| Implementation support and resources provided | UNESCO is developing methodologies for ethical impact assessment (EIA) and readiness assessment to support Member States. Includes the Readiness Assessment Methodology (RAM) and the Ethical Impact Assessment (EIA). | Practical frameworks for AI-developing and deploying companies; tools to design and deploy AI inclusively; 'Guidelines for AI and Shared Prosperity' which includes a job impact assessment tool, responsible practices, and other resources. |
| Monitoring and evaluation mechanisms | Directs Member States to monitor and evaluate policies, programs, and mechanisms related to AI ethics using a combination of quantitative and qualitative approaches. UNESCO can contribute by developing methodologies for ethical impact assessment, readiness assessment, and policy evaluation. | Ways to measure progress toward specific objectives that ensure AI development and deployment is inclusive by design; developing practical frameworks and policies that companies can adopt; shared prosperity targets track the impact on labor income (job availability and distribution, compensation, and worker well-being). |
| Stakeholder engagement process | Emphasizes the involvement of diverse stakeholders, including ethicists, social scientists, technologists, policymakers, and civil society organizations. Encourages Member States to engage all stakeholders, including business enterprises, research organizations, and academic institutions. | Input from workers, thinkers from academia, the AI industry, civil society, and the public; a multi-stakeholder steering committee guides the initiative; involves a multi-disciplinary body. |
| Transparency and accountability measures | Advances essential principles such as transparency, accountability, and the rule of law online. Emphasizes the need for enforceable transparency protocols and ensuring that ethical and legal responsibility can be attributed to AI actors. Appropriate oversight, impact assessment, audit, and due diligence mechanisms are to be developed to ensure accountability for AI systems. | Public disclosure measures detailing how AI systems work, including algorithmic decision-making processes, data management, and efforts to ensure fairness and accountability. |
| Focus on specific ethical concerns | Addresses potential biases in AI systems, ensuring fairness and non-discrimination. Also focuses on ensuring the privacy and security of data used in AI systems, with strong data protection frameworks and accountability for safeguarding personal information. | Ensuring AI progress expands economic prospects, particularly for workers with limited educational opportunities; addressing the potential for AI to increase inequality and negatively affect low- and middle-income countries; aims to prevent biases and discrimination in AI decision-making processes; considers privacy, security, and the prevention of data misuse and misinterpretation. |
| Alignment with international standards and frameworks | Grounded in international law, particularly human rights law. Aligns with the United Nations Guiding Principles on Business and Human Rights. | Aligns with the G20 AI Principles, which call for an evidence-based policy approach and responsible AI stewardship. |
| Adaptability to different cultural contexts | Recognizes the diversity of ethical orientations and cultures around the world. Emphasizes the need for context-sensitive, long-term perspectives on the social embedding of AI technologies. | Global in scope, recognizing that AI impacts the economic future worldwide; recognizes that government policy is usually focused on the welfare of a single nation. |
| Accessibility and ease of understanding | Promotes public understanding of AI and data through open and accessible education, civic engagement, digital skills, and AI ethics training, media and information literacy. | Provides a 10-step guide to responsible AI adoption in newsrooms. |
| Evidence of impact and effectiveness | UNESCO is focused on implementing the Recommendation and is piloting tools, establishing regional roundtables for peer learning, and developing networks of partners. The creation of a Global Observatory on AI Ethics will serve as a one-stop shop for analysis on the ethical development and use of AI. | 'Redesigning AI for Shared Prosperity: An Agenda' has been cited by a leading AI company and academics. |
| Cost and funding model | Member States should ensure sufficient public funding to support programs addressing the challenges identified, including upskilling and reskilling. | Cost of implementation not detailed; receives financial support from organizations like the Ford Foundation; Korinek acknowledges financial support from the Partnership on AI (PAI)'s Shared Prosperity Initiative (SPI). |
| Price | Not available | Not available |
| Ratings | Not available | Not available |