New research explores AI’s effects on governance, adoption, and work

Authors: James Lee, Rosie Hood

Institutional credit: LinkedIn Economic Graph Research Institute

Abstract:
This paper brings together recent LinkedIn Economic Graph research on AI, labour markets and workforce adaptation, alongside new analysis prepared for the LSE Global Forum for AI and the Social Sciences. Drawing on data from more than 1.3 billion LinkedIn members worldwide, it examines how workers, firms and labour markets are responding as AI systems become increasingly capable and widely adopted.

Across major economies, we find little evidence of an economy-wide hiring shock attributable to AI. Instead, the data point to uneven adaptation, including weaker entry-level hiring in AI-augmented occupations, new technology-enabled roles, growing entrepreneurial activity and the spread of AI capability across firms and sectors. New analysis of cybersecurity and assurance highlights a wider challenge: trusted adoption depends on organisations combining AI expertise with governance, security, assurance and domain knowledge. As AI systems become increasingly integrated into workplaces, the opportunity is not simply to automate tasks, but to create new pathways into work, augment human expertise and support the responsible deployment of increasingly capable technologies.

Read the paper

Authors: Faidat Abdullahi, Ayantola Alayande, Fola Adeleke

Institutional credit: Global Center on AI Governance, African Observatory on Responsible AI, IDRC Canada, FCDO

Abstract:
Africa is in the midst of a wave of AI governance activity. Governments across the continent are setting out how they intend to develop and deploy AI, and what they expect it to contribute to economic growth and development. Sixteen countries have published national AI strategies, and several more are drafting them. But the substance behind these commitments varies widely. Some strategies include implementation plans, institutional arrangements, and measurable targets; others set out aspirations without detailing how they will be met. This report analyses AI strategies across the African continent. The reportcprovides a detailed account of national priorities in AfricancAI strategies, and assesses how well those priorities are translated into substantive ccommitments ready for implementation and grounded in African realities The findings are intended for policymakers, technology companies, investors, and academic researchers working toward robust and inclusive AI development in Africa.

Read the paper

Authors: Christophe Combemale, Lindsey Lacey, Dustin Ferrone, Ed Teather

Institutional credit: The AI Adoption Initiative, Carnegie Mellon University

Abstract:
AI diffusion depends not only on frontier innovators and widespread user skills, but on the workers and capabilities that translate AI into use. Building on the AI Labor Stack, this paper develops a draft taxonomy of “facilitators” – the missing middle between AI capability and economy-wide adoption. It distinguishes three domains of facilitation: social, technical, and infrastructural, spanning roles from organisational change and governance to AI deployment, data systems, and physical infrastructure. The paper also introduces a “cycle of facilitation”, showing how preconditions, organisational structures, AI applications, and feedback interact to support successful adoption. Using a novel taxonomy of AI-related job postings, it illustrates how firms are building facilitator capacity and how those patterns can inform national workforce and adoption strategies. The framework aims to provide a foundation for measuring facilitator supply and demand, identifying capability gaps, and designing policies that support both faster AI diffusion and new employment opportunities.

Read the paper

Editors: Silvia Rainone, Wouter Zwysen

Institutional credit: European Trade Union Institute

Abstract:
How is collective bargaining across Europe responding to the growing use of AI and algorithmic management (AM) at work? This volume examines the evolving relationship between legal frameworks and collective bargaining through an analysis of eight countries: Belgium, Denmark, France, Germany, Italy, the Netherlands, Poland and Spain. It reveals a diverse landscape of regulatory experimentation, in which the emergence of agreements on AI and AM appears to be shaped by a combination of features of national legal and industrial relations systems, including the presence of ad-hoc legislation granting specific participatory rights.

Yet the analysis also reveals the limits of this experimentation. Collective agreements continue to rely largely on established governance paradigms, including transparency and data protection, information and consultation, training, and restructuring safeguards. Crucially, provisions addressing the redistribution of productivity gains and economic value generated through AI and AM remain largely absent, as do mechanisms allocating accountability to corporate actors that, despite not being contractual employers, control the technological infrastructures through which labour processes are increasingly governed.

Publication date: October 2026

Find out more