Between hype and implementation: How GenAI can really benefit public administration
What can Generative AI Do for Public Administration – and where are its limits? This was the central question of our captain Dr Clemens Ammann’s keynote at this year’s Swiss eGovernment Forum in Bern. As a platform for innovation and dialogue, the forum brings together decision-makers from federal, cantonal, and municipal levels. It was all the more important for him not only to speak about the potential of the technology, but also about its prerequisites, risks, and strategic significance for the public sector.
Between Poetry, Bureaucracy, and Strategic Integration
Clemens began with a small experiment: could the audience tell whether a poem was written by a human or by AI? Most could not – a moment that clearly illustrated how creative and convincing GenAI has already become.
But this is where the real discussion begins: What does it mean when generative AI is used not just in creative but also in administrative processes? Clemens shared examples from literature, meme culture, and music – alongside concrete use cases in public administration, from digital assistants to automated contract generation. The key question remains: How do we deploy GenAI in a way that is strategically meaningful – and not just technologically impressive?
Six Hypotheses – and Many Aha Moments
The core of the keynote critically examined six common assumptions about GenAI. For example:
- “We need an AI-first strategy” – but without organisational anchoring and alignment with overarching goals and business strategy, such a strategy often lacks impact.
- “More AI means less bureaucracy” – yet inefficient bureaucracy can just as easily be automated, potentially amplifying the problem.
- “Proprietary data is a must” – often, clever prompting and the use of existing models can already achieve a great deal.
These insights struck a chord. Many administrations have pilot projects underway, but often lack a shared vision, proper governance, or structural integration. This is where the real work begins. Ideas are plentiful – but without structures, responsibilities, and learning spaces, they often get lost in the daily grind.
Success Factor Organisation: GenAI Is Also a Cultural Shift
The greatest challenge is not the technology – it’s the organisation. How can we engage employees, build skills, and foster trust? How do we avoid unregulated experimentation – or the illusion that AI must be deployed everywhere immediately?
A metaphor that resonated well with the audience:
Digitalisation was like a bicycle – it made us faster, but we still had to pedal ourselves.
GenAI is like a car – faster and more convenient. But not every route is suitable for it, and in some places, the roads aren't even built yet. And those who rely too heavily on the car risk losing certain abilities.
That’s why a smart interplay is needed: technological openness, organisational maturity, and strategic reflection. At Flybridge, we advocate for a gradual, thoughtful integration – through sprints, clear guidelines, AI ambassadors, and, above all, shared learning.
Conclusion: It’s Not the Technology That Decides – It’s How We Use It
GenAI is not a self-starter. Technology alone does not drive transformation – what matters is how we apply it: context-sensitively, solution-oriented, and collaboratively. If we see GenAI as a tool for augmentation rather than full automation, if we design it in a human-centred way and think about it strategically, it can create real value.
More information about the event:
- Ammann, C.L. (2025, March 28). Contribution to the Swiss eGovernment Forum on the use of generative AI in public administrations. Mentioned in: Jaun, R., & jor. View of the GenAI laboratories of public administrations. Web week. https://www.netzwoche.ch/news/2025-03-28/blick-in-die-genai-labore-oeffentlicher-verwaltungen
- Ammann, C. (2025, March 26). GenAI in administration: Opportunities & success factors [practice presentation]. Swiss eGovernment Forum 2025, Bernexpo, Bern, Switzerland.
