2025年人工智能与劳动力发展报告（英）

AI & WORKFORCE DEVELOPMENT Building a Secure, Equitable Future for Tech WorkersNovember 20252CONTENTSExecutive Summary 3Our recommendations for the future 4Introduction 5How AI is impacting the tech job market 8Tech Industry Commitments to AI Upskilling 12Pledging resources to training initiatives 12Online courses and bootcamps 13The outlook for industry-led upskilling 14Public Policy Approaches to AI and Labor  16Federal level 16State level 17The outlook for public policy approaches 19Labor Union-Led Efforts in Training and Worker Protection 22Collective bargaining 23Collaboration with colleges 24Union and employer partnerships 25The outlook for labor union-led efforts 26Conclusion 28Acknowledgements 30Citations 3132025 saw a wave of press coverage and headlines warning of major shifts in our lives at work. Policymakers and working people are feeling uncertain about the future and are looking for appropriate policy responses. Previous cycles of labor disruption have shown that a sole focus on reskilling, retraining, and workforce development has been far from sufficient to address the impact of technology-induced shifts in the labor market.Workforce training commitments alone are an insufficient response from tech companies and policymakers to the current and unfolding impact of emerging technology on workers.Workforce development programs can be useful resources as jobs change, but they need to be paired with broader economic policies to ensure that tech’s evolution benefits everyone. Unless we take seriously the question of the power differential between employers and workers, any benefits of AI technology breakthroughs will continue to accrue to the rich and powerful. We share examples and lessons from three main approaches to workforce development: tech industry commitments to AI upskilling; public policy approaches to AI and labor; and labor union-led efforts in training and worker protection. These examples lead us to three significant conclusions regarding the AI-related workforce development initiatives that have been announced in the last three years. • First, industry-led initiatives and, to a lesser extent, public policy initiatives, rarely design training programs based on what workers want and typically do not include mechanisms to integrate worker voice into program development. EXECUTIVE SUMMARY4• Second, absent the leverage of worker power through collectively bargained agreements or community benefits agreements, industry commitments are typically vague and highly opaque—both before programs are started and after they are completed. There is little data to be found on the outcomes of corporate training programs, whereas there are numerous studies that have produced robust evidence on publicly-funded programs. • Third, programs to ensure that the skills of the labor force match the job market are only one small piece of the picture when policymakers are considering how to ensure that working people are protected as technology