UK Public Sector Leads Europe In GenAI Trials At 75% Adoption
The Capgemini Research Institute's 2025 survey of 350 government agencies indicates that 64% have initiated pilots or early rollouts of generative AI. Sectors such as defense, health, and security show significant activity, with 82%, 75%, and 70%…
The Capgemini Research Institute's 2025 survey of 350 government agencies indicates that 64% have initiated pilots or early rollouts of generative AI. Sectors such as defense, health, and security show significant activity, with 82%, 75%, and 70% respectively, already testing AI initiatives. This is attributed to larger budgets and critical missions.
Capgemini reports that 90% of agencies aim to explore, pilot, or implement autonomous agents within three years. These systems are expected to manage tasks like drafting letters and routing complaints.
Despite high interest, only 21% of agencies have operational AI services available to citizens, as challenges arise beyond the pilot phase.
Capgemini classifies agencies into four categories. Approximately one-third are "frontrunners," successfully integrating modern tools with routine data use, which results in higher returns. Conversely, 67% lag, either due to outdated infrastructure or insufficient training.
Infrastructure is a significant challenge, with 77% of executives citing inadequate hardware for AI workloads. Budget constraints affect 65% of agencies, preventing major upgrades. Governance issues also persist, with more than half of respondents indicating inconsistent rules for privacy and audits.
The Capgemini Research Institute's 2025 survey of 350 government agencies indicates that 64% have initiated pilots or early rollouts of generative AI.
Capgemini identifies seven components of data maturity, ranging from record identification to insight activation. Less than 25% of agencies rate themselves as highly mature in any area. Only 10% operate scalable cloud platforms, and 12% incorporate insights into daily operations.
Skill gaps exacerbate these issues, as only 29% conduct job-based training for most employees, and 38% train staff to effectively communicate data insights to management.
Challenges in Data Sharing and Energy Usage
Culture impedes change, as risk-averse routines prevail and staff fear errors in handling personal records. Sweden's public employment service emphasizes the need for clear goals and open data discussions to foster trust.
Data sharing is limited, with only 35% of agencies having implemented sharing schemes, and just 8% achieving full coverage. This is linked to trust issues between departments. Energy consumption is another concern, with 75% of agencies worried about the power demands of large models, though only 59% have measures in place to address it.
Capgemini's report highlights that 75% of UK agencies are experimenting with generative AI, leading their European counterparts. This is supported by early cloud adoption and a governmental push for "AI-ready" services. Additionally, 90% of UK departments plan to pilot agentic AI, focusing on applications like chatbots for tax inquiries and tools for sorting social service cases.
Data protection and sovereignty concerns are significant, affecting 78% of agencies. Cost pressures impact 68%, and 75% cite the power and water costs of large models as issues. Only 21% of agencies possess sufficient clean data for new systems, making secure data sharing a priority for the UK government.
Leadership roles are evolving, with Chief Data Officers present in 60% of UK bodies and Chief AI Officers in 27%, with further recruitment anticipated.
Based on reporting by techround.co.uk.