The adoption of artificial intelligence (AI) in the public sector is now reaching a stage where, drawing on the experience of early pilots and adoptions, EU public administrations are starting to face the challenges of implementing AI solutions. In response, this study investigates AI adoption in the public sector with a twofold goal: - Add evidence to the existing body of knowledge to have a better understanding of the dynamics underlying AI adoption in the EU. We do this by providing quantitative (survey) insights into AI readiness and adoption in the public sector, across different country contexts. By offering a picture of the status of AI adoption and readiness in public administrations, we identify the main challenges and drivers of AI adoption, which are required for ensuring AI's trustworthy use. - Define recommendations for managers in the public sector and public administrations. Based on the insights from the first aim, we formulate ways forward to inform policymakers. We surveyed 576 public managers in seven countries: Germany, Spain, France, the Netherlands, Austria, Poland and Sweden. The sample was diverse in age, job level, organisation size and geographical origin. We asked each of them about the level of AI adoption in their organisation. This was measured in two ways: we asked specifically about the extent to which they thought that their organisation had implemented AI projects in service delivery, internal operations and policy decision-making. Next, we asked about the exact number of projects that were either planned or implemented, with the response options of 0, 1, 2-5 or more than 5. Building on the latest scientific insights, we look at what combination of technological, organisational, environmental and individual-level factors contributes to AI adoption. Based on our research, we have three key conclusions: 1. AI adoption is no longer a promise; it is a reality, in particular for service delivery and internal operations. 2. Soft factors and in-house expertise are important internal factors for AI adoption. 3. Citizen needs are an important external factor for AI adoption.
