Artificial Intelligence for Digital Public ServicesAI4DPS 2025

The digitization of public services is a worldwide endeavour. Progress and success of it are hampered by a wide array of legislative and political conditions, including

• the fragmentation of systems in public administrations, the lack of standardization and reuse of data and software, and the autonomy of administrative structures, which hinder coordinated actions;

• crosscutting concerns such as data protection, cybersecurity, and accessibility;

• rapid technological shifts, such as the migration into the cloud and the integration of chatbots in the communication with the citizen; and

• an increasing fear of technological dependency (vendor lock-in) and concomitant loss of sovereignty.

The availability of Generative AI (GenAI) tools already has a profound impact on software development: boilerplate code need no longer be searched, copied, and adapted from various resources, but can be prompted to suit specific needs. On a larger scale, GenAI can analyse complex, fragmented IT landscapes to generate high-level architectural blueprints and integration strategies, aiding in identifying common patterns and proposing standardized interfaces that bridge the gaps between autonomous administrative systems, thus fostering a unified IT strategy. Last but not least, GenAI may be involved directly in helping the citizen with accessing public services, both by identifying the processes that address individual needs and by lowering technological barriers to their use.

This inaugural workshop aims to position GenAI in the transformation of public services by bringing together

• Research on how GenAI can drive the digital transformation of the public sector by fundamentally improving inefficient, document-centric processes. This includes insights from scoping literature reviews, case studies, and real-world implementations of GenAI in public administration.

• Identification of the grand challenges of integrating GenAI into public services, including data security, ethical concerns, transparency, accountability, and public trust.

• Necessary skills and awareness in the public sector for responsible GenAI development and its diffusion in digital public services. What are the regulatory constraints? Are there loopholes to be closed?

• Success stories: Where and how has GenAI been successfully used in the development of software for the public sector? Where and how has GenAI been turned directly into a service? Where and how has GenAI been employed in the dialogue with citizens?

• Pitfalls: What are common misassumptions? What are the false promises of GenAI? Which drawbacks need to be addressed proactively in the application of GenAI in the three usage scenarios outlined above?

• Political consulting: How to inform politics about the real, sustainable potential of GenAI in a publicly responsible way? How to debunk overselling and lobbying of GenAI with resonation in the media?

Accepted Papers

Call for Contributions

We welcome the submission of position papers of max 2 pages, with content suitable to fuel the discussion during the workshop. All submissions will undergo a light (informal) reviewing process conducted by the Organizing Committee. While every accepted submission will be given presentation time, the focus of the workshop is on discussion.

Submission Guidelines

• Submitted position papers shall include: title, motivational statement, and the intended contribution to the discussion.

• Contributions can be in English or German.

• Submissions must be made via EasyChair, using https://easychair.org/conferences/?conf=staf2025.

• All contributions require at least a one-day registration (as does of course attendance).

• Workshop proceedings will be published electronically, as a collation of the submitted position papers.