Global MPM Insight

Latest global trends in public personnel administration : a view from the OECD Global MPM Insight Vol.5 to enable adoption. This shift toward integrative, future- oriented leadership is essential for embedding AI into core business systems and ensuring that workforce transformation is intentional rather than incidental to technological change. Senior leaders are also central to the strategic workforce choices that determine how AI reshapes roles, sk i l l s, and career pathways across publ i c services. This places responsibility on leaders to update competency frameworks, anticipate areas of declining and growing demand, and invest in structured reskilling pathways so that employees can transition into more complex, AI-augmented work. By shaping roles proactively and ensuring that institutional knowledge is retained and redeployed, senior leaders act as stewards of a long-term workforce transition in which AI becomes a complement to public-sector expertise rather than a substitute. Artificial Intelligence and transformation Across the OECD, the capabilities of generative AI, machine learning, and predictive analytics are advancing rapidly. As these and other digital tools become increasingly powerful and accessible, their potential to speed up processing, automate time- consuming tasks, is now widely recognised. But AI is not just about shaving time off discrete processes. Its longer-term potential is as a tool to help governments do what they have been trying to do for many years with limited success – speed up decision- making, develop more flexible and efficient organisations, and accelerate delivery of more user-friendly public services. Yet despite this technological progress, implementation across government remains patchy: many promising initiatives are still confined to pilots, proofs-of-concept, or narrowly defined use cases. Large-scale deployment where AI meaningfully reshapes workflows and service delivery remains mostly the exception rather than the norm. 4) For senior public-service leaders, steering transformation in an AI-enabled administration requires far more than introducing new tools: it calls for reshaping the organisational conditions that allow AI to create value. The types of digital capabilities are not confined to technical specialists – they also depend on leaders who can connect new technologies with day-to-day operations, long-term planning, and how teams and services are organised. For example, Estonia’s virtual assistant illustrates how public service leaders can connect emerging technologies with daily operations, long-term planning, and organisational arrangements. 5) Public service leaders coordinated multiple agencies to deploy a shared, citizen-facing tool built on the country’s existing digital infrastructure and a modular, open-source architecture. This required managing procurement choices, interoperability constraints, and role definitions shaped by data-protection requirements. The system combines generative AI with retrieval and routing components so agencies can provide consistent information through a common interface. Early evidence shows uptake across more than 16 agencies and reported improvements in accuracy, offering insight into how coordinated leadership can support wider adoption of such tools. This example raises another skill set – to design quality trials and experiments that are able to evaluate positive, and potentially negative impacts of the technology, as a precursor for potentially broader roll-out. In this light, effective leadership means re-framing the approach to risk, enabling responsible experimentation with new technologies, and redesigning workflows so that human judgement and AI systems are integrated coherently rather than bolted onto legacy processes. Leaders need enough AI literacy to identify where transformation is possible, set expectations for safe and ethical use, and align incentives, resources, and cross-functional teams Ageing labour markets, skills, and workforce expectations Ageing public workforces are now a defining feature across many OECD countries. Across the OECD, the share of employees aged 55 and over is higher in central and federal administrations than in the wider labour market, and between 2015 and 2020 this proportion rose by 1.2 percentage points to reach 26%. 6) Central government bodies are typically older than the labour force overall, and a substantial cohort is expected to retire in the coming years. This raises not only concerns about long- term workforce capability, continuity, and institutional memory, but also questions about how leaders prepare their organisations for a period of large-scale demographic transition alongside measures to downsize the public service. In this context, senior leaders must anticipate where capability risks will emerge, how knowledge can be transferred before it is lost, and how career development models can change so that employees can continue to grow and contribute meaningfully at every stage of longer working lives. This requires recognising that forthcoming retirements can either reinforce existing skills gaps if unmanaged, or provide a strategic opportunity to renew skills, diversify entry pathways, and redesign roles to better support transformation. Seizing this moment is an opportunity for renewal of the public workforce to better support transformation. As roles in the public service evolve and retirements reshape the age and skill profiles in ministries, administrations can rethink career pathways, expand mobility programmes, and shift towards skills-based job structures that better match future service needs. These shifts are essential for managing downsizing pressures without unintended losses of critical capability. 4) OECD (2025), Governing with Artificial Intelligence: The State of Play and Way Forward in Core Government Functions, OECD Publishing, Paris, https://doi.org/10.1787/795de142-en. 5) https://www.ria.ee/en/state-information-system/personal-services/burokratt 6) OECD (2025), Government at a Glance 2025, OECD Publishing, Paris, https://doi.org/10.1787/0efd0bcd-en. 22 23