Author: Federal Office of Information Technology and Telecommunications (BIT)
Source: https://www.bit.admin.ch/de/bit-stories-de-83
Publication Date: December 9, 2025
Reading Time: approx. 5 minutes
Executive Summary
The Swiss federal administration uses AI strategically and responsibly – not as an end in itself, but as a tool for increasing efficiency and improving citizen experience. The BIT customer event on May 21, 2025 with 380 participants demonstrated three concrete success stories: the Alva chatbot (Basel-Stadt), the internal BIT AI Assistant, and research-based recommendations for action. The key lies in clear objectives, high data quality, transparency, and preserving human responsibility.
Critical Guiding Questions (Liberal-journalistic)
Freedom & Transparency: Are citizens and employees adequately informed that they are interacting with AI systems, and do they retain control over their data?
Responsibility: Who bears liability if AI systems make erroneous administrative decisions – the authority, the developers, or the user?
Innovation vs. Security: How is the balance between experimental learning and the protection of sensitive administrative data ensured?
Democratic Legitimation: Are AI decisions in administration sufficiently reviewed by parliament and the public?
Equal Opportunity: Do all population groups benefit equally from improved digital services, or are new digital divides emerging?
Scenario Analysis: Future Perspectives
| Time Horizon | Expected Development |
|---|---|
| Short-term (1 year) | Rollout of AI services for BIT customers (planned for 2026); Alva will be expanded with conversational capabilities; further pilot projects start in other cantons. |
| Medium-term (5 years) | AI becomes standard tool in administrative processes; citizens expect AI-supported chatbots as the norm; data protection and compliance become competitive advantages. |
| Long-term (10–20 years) | AI could fundamentally change administrative workflows; Risk: automation leads to job losses without retraining programs; Opportunity: employees focus on strategic tasks. |
Main Summary
Core Topic & Context
The Swiss federal government positions itself as a responsible AI user in public administration. After the ChatGPT hype (since November 2022), it is no longer about technology for its own sake, but about concrete added value for citizens and efficiency gains for authorities. The 2025 BIT customer event documents this pragmatic approach through three practical examples.
Key Facts & Figures
- 380 participants (in-person and online) at the BIT Stories 2025 customer event
- Alva chatbot: Based exclusively on publicly accessible content; uses Retrieval-Augmented Generation (RAG)
- BIT AI Assistant: Available internally since 2025; rollout for customers planned for 2026
- Language support: Alva works surprisingly well in multiple languages (Details ⚠️ not specified)
- User behavior: Many users prefer search terms over chatbot dialogue (⚠️ no figures provided)
Stakeholders & Affected Parties
| Winners | Losers / Risks |
|---|---|
| Citizens: Better access to public information | Data protection skeptics: Concerns about data processing |
| Administrative employees: Time savings on repetitive tasks | Low-skilled administrative staff: Automation risk |
| Authority leadership: Efficiency gains, modernization image | Small cantons: Resource shortage for AI projects |
| Research (EPFL): Knowledge gains from practical projects | Critics: Lack of parliamentary oversight |
Opportunities & Risks
| Opportunities | Risks |
|---|---|
| Faster access to administrative information for citizens | Incorrect AI responses influence administrative decisions |
| Time savings for employees → higher job satisfaction | Data leaks due to insufficient security |
| Standardization of processes across authorities | "Black box" decisions without traceability |
| Attractiveness for IT talent; innovation in public sector | Dependence on external tech providers (⚠️ not discussed) |
| Multilingual services with less effort | Digital exclusion for non-digital population groups |
Action Relevance
For decision-makers:
- Act now: Develop AI strategy with clear success criteria (focus on benefits, not just technology)
- Create transparency: Publicly communicate where AI is used and why
- Strengthen responsibility: Train employees to critically review AI outputs
- Manage risks: Data protection and security as prerequisites, not afterthoughts
- Parliamentary oversight: Regular reporting on AI projects and their impacts
Quality Assurance & Fact-Checking
- [x] Central statements verified (Alva project, BIT AI Assistant, EPFL research)
- [x] Unconfirmed information marked with ⚠️ (language support, user figures)
- [x] Official BIT source used
- [x] No political one-sidedness detected – objective, pragmatic tone
Note: The article is a self-presentation by BIT. Critical opposing voices (data protection advocates, unions, parliamentarians) are absent.
Supplementary Research
- Federal Council Report "AI Strategy Switzerland" (2024) – official government line
- Federal Data Protection Commissioner (EDÖB): Statements on AI in administrations
- Parliament: Inquiries and debates on AI use (Suisse-Parlament.ch)
Bibliography
Primary Source:
BIT Stories 2025 – From Idea to Practice: AI Projects in Administration
https://www.bit.admin.ch/de/bit-stories-de-83
Supplementary Sources:
- Alva Basel-Stadt: https://www.bs.ch/alva
- EPFL research on AI in public administration (Amin Kaboli)
- BIT contact: Suela Amin, Head of Content & Event Services
Verification Status: ✓ Facts checked on December 9, 2025 (publication date)
This text was created with the support of Claude Haiku 4.5.
Editorial responsibility: clarus.news | Fact-checking: December 9, 2025