The adoption of Generative AI in EU public administrations: exploring individual behaviours and organisational approaches

Generative artificial intelligence (GenAI) is rapidly reshaping public sector work and accelerating policy debates across Europe. Public administrations are increasingly experimenting with GenAI tools to support document drafting, knowledge management, information processing and service delivery, while simultaneously facing growing challenges related to governance, data protection, organisational readiness and technological sovereignty. Against this background, this report examines how GenAI is being adopted and integrated within European public administrations, through interviews, case studies and a review of the literature. It focuses both on the individual use of the technology by civil servants and on the broader organisational processes required for its strategic assimilation. At the individual level, GenAI adoption is characterised by persistent tensions between empowerment and dependency, innovation and control, and experimentation and regulation. Civil servants increasingly perceive GenAI as a valuable cognitive assistant capable of enhancing productivity, supporting decision-making and improving access to organisational knowledge. At the same time, concerns have emerged regarding reliability, over-reliance, skill erosion, and the widespread use of external commercial tools outside formal organisational channels. At the organisational level, GenAI adoption requires more than the deployment of technological solutions. The assimilation of GenAI encompasses a careful consideration of several key factors: the availability of secure and user-friendly environments, governance capabilities, data and technological infrastructures, workforce skills, and mechanisms for experimentation, monitoring and continuous adaptation. Different assimilation strategies are possible, depending on the type of value administrations aim to generate through GenAI use. Lighter approaches may be sufficient for operational support, while more advanced forms of cognitive augmentation and human–AI collaboration require deeper organisational integration and more strategic governance arrangements. The value-based framework proposed in this report is intended to support reflection on this topic. Building on these findings, administrations and EU Member States should reflect on how to best harness the opportunities offered by GenAI. The report proposes a set of operational policy recommendations to support the responsible, secure and scalable adoption of GenAI in the public sector. Overall, it contributes to the emerging debate on how European public administrations can move from fragmented experimentation towards a more strategic and governed integration of GenAI into administrative processes and public service delivery.