The Landesverwaltung's strategy focuses on responsibility, human centrality, transparency and efficiency to improve services to citizens

Il Liechtenstein chose to face theartificial intelligence in public administration not as a simple technological equipment, but as a theme of organization, responsibility and institutional trust. In the session of Tuesday 14 April 2026, the Government adopted the strategy for the use of AI in Liechtensteinische Landesverwaltung, the Principality's national public administration. Its stated objective is to contribute to a more modern public administration, closer to citizens and sustainably oriented towards the well-being of society.

The decision comes at a time when many European administrations are evaluating how to use generative systems, document automation tools, semantic search engines and predictive applications without compromising rights, procedural fairness and human controlFor a small-sized state, but with an administration called upon to guarantee high standards of service, the issue is not just about the digitalization of processes: it concerns the ability to define internal rules before the adoption of new tools occurs in a fragmented way.

This is precisely the most important point of the strategy. The government does not present artificial intelligence as a shortcut to replace administrative functions, but rather as a tool to be prudently integrated into the working life of the Landesverwaltung. The official source outlines four guiding principles: responsible use, anthropocentric approach, transparency towards employees and the population, orientation towards efficiency and innovationThese are principles which, if translated into operational practices, can impact the selection of tools, staff training, data management and control criteria. assisted decisions.

The Principality of Liechtenstein's policy document on artificial intelligence
The Principality of Liechtenstein's policy document on artificial intelligence (short version)

Four principles to avoid a disorderly adoption

The first principle is the responsibilityThe strategy establishes that AI integration must comply with legal requirements and high ethical standards. In the public sector, this step is crucial, because the administration does not operate like a private company optimizing a commercial service: it manages practices, rights, sensitive information, procedures, and institutional interactions. Any use of algorithms or language models must therefore be compatible with legality, proportionality and administrative correctness.

The second principle is the centrality of the personThe source specifies that artificial intelligence tools must support employees without transferring responsibility for actions to machines. This is an important formulation: it does not promise an automated public administration, but rather a assisted public administrationIn organizational terms, this means that the official remains responsible for the outcome, verification, and final decision, while the digital system can help reduce repetitive workloads, speed up internal searches, or improve document preparation.

The third principle concerns the transparencyThe use of AI must be communicated in a way that is understandable to both employees and the general public. This choice has political and administrative implications: if citizens and users don't understand when and how an automated tool enters a process, the risk is to foster mistrust, even when its application is limited and controlled. Transparency therefore becomes a key factor. element of legitimacy, not a subsequent communication detail.

The fourth principle unites efficiency and innovationThe strategy aims to improve the quality of services and to encourage the evolution of the Landesverwaltung. The message is measured: AI is not presented as an immediate revolution, but as a factor capable of accompanying the administrative modernizationIn this sense, potential applications could include text processing, request classification, internal knowledge support, information flow management, or assistance in document preparation, always within defined boundaries.

“This structure ensures that artificial intelligence is used holistically and that the further development of the national administration takes place on a broad basis,”

underlined the Regierungschefin Brigitte Hass, indicating in the

“combination of people, organization, governance and infrastructure is the perimeter within which to transform a technological choice into a controllable administrative path".

People, governance and infrastructure as operational axes

The strategy identifies four fields of action: person, organization, governance and infrastructureThis distinction is significant because it avoids reducing AI to a software problem. In many public organizations, in fact, the most common mistake is to introduce digital tools without adapting them. roles, skills, responsibilities and technical architecturesThe Liechtenstein government, on the other hand, appears to want to include innovation within a broader management framework.

Field "person” calls for the need to train staff and build a culture of conscious useGenerative AI, in particular, can produce fluent but not always accurate texts, can summarize documents incompletely, and can return seemingly plausible but unverified results. In a public administration, where even a linguistic nuance can have operational consequences, training cannot be limited to technical familiarity: it must include critical thinking, source checking, and an understanding of the tools' limitations.

Field "organization" It is instead about the ability to adapt processes and responsibilities. AI can generate benefits only when it is inserted into clear workflowsWho can use a given tool, for what tasks, with what data, with what controls, and with what exclusions. Without this definition, technology risks creating informal practices, disparities between departments, and uncertainty about outcomes. The strategy, in fact, requires an organizational concept that defines the roles, responsibilities, and procedures of the relevant offices.

Field “governance” It's the most delicate. In a public administration, AI governance must establish approval, monitoring, audit and update criteriaIt must also distinguish between low-risk tools, such as internal editorial assistants, and more sensitive uses, where automated processing can impact the relationship with citizens. Governance is not a brake on innovation: it is the condition for making it defensible, replicable, and consistent with the public mandate.

Field "infrastructure" introduces the topic of technical foundation. The adoption of AI requires secure environments, accurate data management, access policies, integration with existing systems, and the ability to protect administrative information. For a public entity, the choice between external solutions, controlled environments, proprietary tools, or integrated platforms is not neutral. It impacts the data sovereignty, on maintenance costs, on dependence on suppliers and on the possibility of verifying the behaviour of the systems.

The value of public AI depends on the limits we set

One of the most interesting parts of the strategy is the choice not to separate innovation and ethicsThe source emphasizes compliance with regulations and high standards, but also insists that employees never surrender their responsibility for their actions. This approach reflects a widespread awareness in the public sector: AI is useful when it provides support, not when it obscures decision-making. The quality of administration depends not only on the speed with which a matter is processed, but on the ability to explain the process that led to a result.

According to industry analysts, the main issue for administrations is not to establish whether or not to use artificial intelligence, but to define which functions can be assisted without altering the institutional responsibility.

AI can generate value when it reduces repetitive work, helps organize knowledge, and improves departments' responsiveness, but it becomes problematic when it enters decision-making areas without clear criteria, traceability, and trained personnel to verify its outputs. Public administration must therefore proceed with measurable use cases, documented controls, and a clear distinction between technical assistance and administrative decisions.

experts observe in summary public digital transformation.

From this perspective, Liechtenstein's strategy can be seen as an attempt to anticipate the typical critical issues of AI adoption. Transparency towards the population, for example, is not just about external communication. It can become a design criterion: informing whether a text has been prepared with algorithmic support, clarifying when a system helps sort a request, explaining which data is not used or which tasks are excluded from automation. Trust often arises from these procedural details.

Efficiency, too, must be interpreted with caution. In the public sector, speeding up doesn't necessarily mean simplifying. A faster but less verifiable process can generate more problems than it solves. Administrative innovation therefore requires: qualitative indicatorsAccuracy, consistency, error reduction, clarity of responses, accessibility of services, reduced burden on offices, and greater comprehensibility for citizens. The strategy does not yet provide a list of operational applications, but it paves the way for defining them within a unified framework.

A dynamic strategy until the 2030 review

The document approved by the Government is described as a dynamic strategyIt has no fixed validity period and will be subject to regular checks and updates. This choice appears consistent with the nature of artificial intelligence, a field in which tools, risks, technical capabilities, and social expectations change rapidly. Establishing a rigid strategy today would risk its rapid obsolescence. Conversely, an updateable document allows objectives and priorities to be adjusted without having to start from scratch each time.

“Goals can be adapted, prioritized, or integrated,”

explained Brigitte Haas.

“In this way we adequately take into account the highly dynamic field of artificial intelligence.”

For this very reason, the strategy does not foresee a fixed duration: a period of 2030 at the latest is planned. fundamental revision, during which the strategic objectives will be reviewed and the fields of action updated.

The 2030 deadline has concrete value. It sets a timeframe within which to assess whether the measures adopted have produced organizational benefits, whether governance has been adequate, and whether the technical infrastructure has supported the evolution of services. It also prevents the strategy from remaining an unverified policy document. In a fast-paced sector, periodic review becomes part of the strategy itself.

For Principality of Liechtenstein, the theme also takes on an institutional positioning dimension. A relatively small public administration can experiment more agile governance models Compared to very large state structures, but at the same time it must avoid excessive technological dependencies. The smaller scale can foster coordination, cross-sector training, and use case monitoring. On the other hand, it requires particular attention in selecting skills and infrastructure, because limited resources can make any poor decisions more difficult.

The strategy of the Landesverwaltung does not therefore announce a spectacular transformation, but a path of gradual and controlled integrationThis is precisely its most interesting innovative element. Artificial intelligence enters the public sector not as an indistinct promise, but as a technology to be governed through principles, roles, processes, and infrastructure. At a time when AI is often portrayed as a universal accelerator, Liechtenstein offers a more sober reading: administrative innovation produces value only when it remains comprehensible, verifiable, and firmly anchored in the human responsibility.

Here are three insights that might interest you:

In Liechtenstein, the innovation of the first corporate state in history
digital-liechtenstein.li: Board reviews strategic plans
Liechtenstein Leads the Way for Blockchain Stamps