Onboarding AI from organisational learning perspective

Sammanfattning

This study explores how knowledge workers in multinational service-sector organisations engage in organisational learning during the adoption of generative and agentic artificial intelligence (AI). Drawing on Ellström’s two logics of learning, the study examines how the perceived features of AI — whether concrete or abstract — trigger distinct sensemaking processes that shape two modes of organisational learning: exploitative (adaptive, feedbackdriven) and explorative (developmental, feedforward-driven). A qualitative case study was conducted within the AI Sweden ecosystem using semi-structured interviews with 24 senior knowledge workers and human resources practitioners from service sector organisations. Thematic analysis, informed by Ellström’s learning framework, was applied to the data.The findings reveal that concrete AI features encourage exploratory and developmental learning through experimentation, collaboration, and the agency of early adopters, while abstract AI characteristics tend to produce adaptive learning shaped by uncertainty and topdown structures. Early adopters emerge as the primary drivers of AI learning across organisations, substituting for absent formal learning structures through informal knowledge sharing. Explorative learning is found to be heavily culture-dependent and remains fragile and unsupported at the organisational level, while exploitative learning is procedural but lacks clear outcome measurement. AI learning has not yet been institutionalised in most organisations studied, it remains concentrated at the individual and team levels, with feedback, evaluation, and coordination mechanisms largely absent.The study contributes to the understanding of AI-supported workplace learning by demonstrating how prospective sensemaking processes mediate the relationship between AI characteristics and organisational learning modes, and by identifying the structural conditions that facilitate or constrain knowledge workers’ competency development during AI integration. Practical implications are drawn for human resources practitioners and or ganisational designers seeking to build sustainable AI learning cultures in knowledge-intensive organisations.