Recognizing the potential for both good and harm in AI is essential to developing systems that benefit everyone. The United States forms the frontier for these AI systems and products, highlighting our commitment to leading the responsible development and deployment of AI globally. This commitment is demonstrated in a recently published book. Global AI Development Handbookwas developed in response to the Biden administration. Executive Order (EO) emphasizing safe, secure, and reliable AI development.
A key government resource and product, the Handbook emphasizes that AI is a global opportunity that will have a profound impact on humanity. Although AI has the potential to benefit society, it also poses risks to those who are most susceptible to the negative effects of AI systems. This handbook marks a step forward in the United States’ efforts to lead in building a global coalition on AI policy rooted in human rights and sustainable development.
Due to the lack of global policy and regulatory consensus regarding AI research and development, the EO, in collaboration with the Department of Energy and the National Science Foundation (NSF), urges the United States Agency for International Development (USAID) and the U.S. Department of State to: I was instructed to give instructions regarding the following. develop Global AI research challenges. The research agenda aimed to ensure the safe and responsible global development of AI through international research collaborations, study the impact of AI on humans, and address labor market changes caused by AI.
PAI recommendations
Partnerships on AI have contributed to the research agenda by addressing challenges such as: Call for public comments To inform the agenda and playbook. The Agenda and Handbook contain a lot of content PAI Key Recommendationswhich includes:
Prioritize a thorough investigation of your data supply chain We map the different types of employment models that data enrichment workers fall into, assess the prevalence of data enrichment employment models, analyze the diverse actors involved across global supply chains, and understand how these models Investigate the impact on
pursue empirical research Investigate whether and to what extent the circumstances and working conditions surrounding the construction of a dataset affect its validity, reliability, and safety.
Consider how to design the market Promote fairer outcomes for those involved in data supply chains.
Investigate the impact on quality of work Focusing on the impact on the labor market, in addition to the impact on employment possibilities and wages.
provide an opportunity It aims at non-academic participation and engagement, with a particular focus on incorporating the perspectives of workers and affected communities.
These recommendations focus on important but often overlooked areas of AI policy. They aim to ensure that AI developments not only advance technology, but also contribute to fairer and safer outcomes for workers and those affected by them. A better understanding of the role of data enrichment workers in the development of these systems will uncover unfair working conditions that can and do impact the quality and reliability of AI systems. You can. We have dedicated the past five years to addressing these challenges at PAI. Outline a path to responsible data enrichment practices For worker representatives, policy makers, auditors and other stakeholders across the supply chain.
recently released draft Vendor contract guidance It helps AI development companies promote responsible practices with downstream stakeholders. transparent template Outline what companies should monitor and report regarding their data enrichment practices. our Data enrichment procurement guidelines Influence how AI practitioners engage and collaborate with data enrichment workers to ensure appropriate working conditions.
Our recommendations suggest that the research agenda follows a broad definition of “labor market impact” and includes examining the impact of AI on employability, wages, and job quality.
Agenda provides a unique opportunity to pursue unexplored research areas identified in previous studies. Impact of AI Informal labor markets, expected AI exposure in low- and middle-income economies, and Mechanisms to strengthen worker voice. Engaging the hundreds of millions of people who perform the critical tasks of cleaning and labeling datasets, including data enrichment workers, in these conversations can go a long way toward improving business outcomes for workers and companies. It’s important. Especially affected workers and communities have expertise and real-world experience that is invaluable in informing research efforts. See PAI Overview. 2023 Multi-Stakeholder Forum on Science, Technology and Innovation for the SDGs by the United Nations Department of Economic and Social Affairs On the importance of a multi-stakeholder approach.
Integrating multi-stakeholder voices into AI
As policymakers at the national and international levels work to manage the development, deployment, and use of AI; Bringing ideas across sectors and disciplines into policy discussions is essential so that solutions work for people, not just businesses.. We encouraged USAID, the US Department of State, the Department of Energy, and the NSF to collaborate with organizations like us in designing and implementing the global research agenda. A multi-stakeholder approach to agenda development ensures that innovative and fresh approaches are reflected in the agenda and resulting deliverables. Partnerships on AI play a critical role in shaping responsible global AI governance by bringing together experts from different fields at the intersection of industry, civil society, government, and academia. . PAI continues to tackle the most important and difficult questions about the future of AI. To stay up to date on our work in this area, please visit Newsletter.
