Abstract
In 2024, three quarters of the Austrian population were already using electronic government services, also known as e-government, with an upward trend. Despite the increasingdigitalization, the provision of e-government services in Austria still requires a considerable amount of manual work on the part of the authorities. Furthermore, citizens often
face the challenge of extracting relevant information from a variety of online services. In
this context, artificial intelligence (AI) systems present a promising avenue for automating e-government services, particularly through technologies such as machine learning
and natural language processing. These technologies have the potential to enhance administrative efficiency, increase user-friendliness, and boost citizen satisfaction.
Despite the adoption of a strategy for the broad deployment of AI in the public
sector, Austria currently lacks a substantial number of AI-based e-government services.
The central thesis is therefore that the current range of Austrian e-government services
can be improved by using AI. The first research question to be addressed was thus:
’Which existing Austrian e-government service that is currently operated without AI is
suitable for implementation using an AI process?’
To answer the research question, e-government services in Austria were collected
and it was determined how AI-based services are offered in both Austria and Europe. A
promising use case in this area was identified as a so-called ’RIS chatbot,’ which could
facilitate access to legal information from the Federal Legal Information System (RIS).
Such a chatbot could enable citizens to ask questions about laws and legal frameworks in
natural language in order to identify and understand relevant legal passages. However,
the challenge lies in designing such a system not only to be efficient but also legally
secure, since incorrect or misleading legal information could have serious consequences.
The second research question to be addressed was therefore as follows: ’How can an
AI-based RIS chatbot be designed, implemented, and evaluated so that it is both legally
secure and complies with information security requirements?’
To answer this second research question, a design science approach was used to
collect requirements from science and practice. These requirements were then cast into
a prototype and evaluated. Future research should focus on the concrete implementation
and further evaluation of the reliability of the results.
| Date of Award | 2025 |
|---|---|
| Original language | German (Austria) |
| Supervisor | Eckehard Hermann (Supervisor) |
Studyprogram
- Information Security Management