Artificial intelligence, which not long ago existed primarily as a collection of pilot projects, is now deeply embedded in public governance: it automates workflows, supports planning and decision-making, and improves the quality of public services with citizens' interests in mind. Notably, international experience is becoming a benchmark for countries that, as part of their digital transformation, seek to boost the efficiency of their administrative apparatus. The Organisation for Economic Co-operation and Development (OECD) report on AI governance compiles more than 200 examples of the technology's application across 11 government structures, marking the gradual transformation of AI into a key component of digital public governance.

Scale of Adoption: What the OECD Statistics Say

In the public sector, AI is used to automate and personalize services, support analysis and forecasting in policy development, detect fraud, and improve the efficiency of civil servants. According to the OECD, roughly 67% of member countries use AI to improve the design and delivery of public services. Governments are prioritizing areas with high transaction volumes and frequent interaction with citizens. The majority of applications (57%) are aimed at automating, simplifying, or personalizing services, followed by decision support, analysis, and forecasting (45%), as well as anomaly detection to enhance accountability (30%).

Estonia: Why "Digital Bureaucracy" Does Not Work

Estonia has deeply integrated AI into its e-government platform through the "Krait" strategy (2019), which rests on four pillars: promoting the use of AI in the public and private sectors, building skills and research capacity, improving the legal framework, and creating a favorable environment. Typical examples include machine learning for detecting anomalies in the national data-exchange system X-Road, algorithms matching labor supply and demand at the Unemployment Insurance Fund, and speech-recognition systems for recording parliamentary sessions. Particular attention deserves the swiftly adopted legislation on algorithmic liability, which defined legal responsibility for automated decision-making systems. Over the past two decades, Estonia has become one of the world's leaders in digitalization — "e-Estonia" — serving more than 1.3 million people.

Singapore and the "Human-Centered" Principle

If Estonia builds its AI on the basis of digital data and institutions, then Singapore is implementing an "AI for the Public Good" strategy, placing the human being at the center of attention. According to the Ministry of Digital Development and Information (MDDI), the National AI Strategy announced in 2019 is aimed at creating economic value, improving citizens' quality of life, and positioning the country as a global leader in AI development and governance. At its core lies the principle that every AI application must deliver tangible benefit to society.

Risks and the "Trustworthy AI" Principle

Alongside the opportunities, countries are also aware of the risks: data bias, insufficient algorithmic transparency, digital inequality between population groups, and the potential erosion of public trust if the technology is not properly regulated. AI adoption is therefore inextricably linked to the concept of "trustworthy AI" and the principles of accountability. Although all countries view AI as a driver of governance efficiency, each chooses its own approach depending on its level of development, institutional capacity, and strategic goals.

Contradictory Data

It is important here to honestly distinguish two levels of information. On the one hand, the OECD's aggregated statistics (67% of member countries, more than 200 examples, shares of 57%/45%/30%) describe a stable overall trend and are based on a single consolidated report. On the other hand, the specific national cases — Estonia and Singapore — demonstrate different degrees of maturity and different priorities: from a mature "electronic" infrastructure with legislation on algorithmic liability to strategies formulated in 2019 that are still in the scaling phase. Thus, the statement about the "deep integration" of AI reflects a direction of development rather than a uniform state across all countries: in some cases, application remains targeted or pilot-based. Moreover, the figures cited in the report were not independently re-verified within the other sources reviewed, which record a similar trend at the level of individual states but do not duplicate the OECD's aggregated indicators.

Lessons for Digital Transformation

The combined experience shows that success is determined less by the presence of technology than by prior process simplification, algorithmic transparency, and the willingness to enshrine liability for automated decisions in law. As former Estonian President Kersti Kaljulaid (2016–2021) noted, the key to success is simplifying processes before digitizing them, so as not to create "digital bureaucracy" but to deliver real benefit. It is precisely this approach — from data and institutions to a human-centered focus — that is becoming a practical benchmark for countries implementing AI in public governance as part of their digital transformation.