As Artificial Intelligence (AI) increasingly reshapes the nature of work, understanding the conditions under which AI-augmentation enhances, rather than undermines, worker motivation is critical. This study integrates Self-Determination Theory (SDT; Deci & Ryan, 2008) and Sociotechnical Systems Theory (STS; Trist & Bamforth, 1951) to examine the psychological and work-design dynamics of AI-augmented roles. SDT emphasises the importance of satisfying autonomy, competence, and relatedness needs for intrinsic work motivation and well-being, including in technology-mediated work contexts (Gagné & Deci, 2005, Gagné et al., 2022). AI may ‘frustrate’ these needs through algorithmic decision-making reducing autonomy (Kellogg et al., 2020), absorption of skilled tasks threatening competence (Raisch & Krakowski, 2021), and reduced human interactions (Glikson & Woolley, 2020). However, when well designed, AI-augmented roles can enrich decision environments, support higher-order skill use, and free time for human connections (Parker & Grote, 2022). Complementing SDT, STS provides work design principles for optimising the joint performance of human and technical subsystems (Cherns, 1976; Clegg, 2000; Parker et al., 2025), clarifying how AI-enabled systems can be configured to support both well-being and effectiveness.
To empirically test this dual-theory framework, we conducted a mixed-methods analysis of 9,272 open-text survey comments from employees of a global mining company. Using keyword screening, SDT-informed need coding, and reflexive thematic analysis (Braun & Clarke, 2006, 2019), comments were classified into AI adoption personas based on need satisfaction patterns. Five personas emerged: Enthusiast (52%), Overwhelmed (26%), Pragmatists (11%), Sceptic (6%), and Disengaged (4%). Each persona was then mapped to a corresponding STS design principle. For example, Enthusiasts reported high satisfaction across all three SDT needs and aligned with STS principles for joint optimisation, whereas Overwhelmed employees expressed competence frustration and highlighted the need for simplified systems, clearer role boundaries and targeted training consistent with minimum critical specification.
These findings show how employee responses to AI augmentation vary greatly across workforce segments. The resulting persona framework provides a practical diagnostic tool for tailoring interventions to address specific need frustrations and system misalignments. This study contributes to the literature by integrating motivational and systems-level theories within the context of AI-enabled work, responding to calls for more integrative approaches to digital transformation (Parker & Grote, 2020; Xu et al., 2023). More broadly, it offers a theoretically-grounded and empirically validated framework for designing AI-augmented work that supports well-being, engagement and sustainable AI adoption.