Just under 70,000 respondents from the public sector, 300 workshops, 60 case studies, 44 local authorities and around three million data points: drawing on an impressive dataset, Prof. Dr Dr Björn Niehaves from the University of Bremen will be presenting the state of AI transformation in Germany’s public administrations at SCCON. He is presenting the findings at the panel discussion ‘The Reality of AI in Germany’s Public Administrations: Where We Stand and What Needs to Be Done Now’.

In a joint research project with the German Association of District Authorities, Niehaves is investigating how future-proof local authorities are when it comes to the use of artificial intelligence. This is because new technologies often take longer to find their way into public administrations. Whether chatbots, virtual assistants or, now, AI agents: “What is being discussed here at SCCON will only reach public administrations on a broad scale with a delay of one to two years,” says the researcher. How successful the transformation proves to be depends primarily on governance and the priorities set.

Three paths to AI transformation

Niehaves and his research team have identified three types of transformation:

• Type 1 – The strategic trailblazer: Transformation processes are steered by strategies, but their impact often remains limited.
• Type 2 – The technological implementer: A strong digital department drives technology and tools forward. However, because organisational commitment is lacking, the rest of the administration does not follow suit.
• Type 3 – The cultural enabler: Empathetic leadership puts people at the centre. The transformation enjoys broad acceptance, but there are difficulties with implementation and scaling.

Three-quarters of the German public administrations studied fall into Type 2. “IT is plugging away, but fails to get the whole organisation on board. And new ideas and technology only make things worse,” explains Niehaves.

But there is also some good news: “The days of mere talk are over everywhere. We are now in the implementation phase. But in doing so, all departments – from HR to the executive board – must be brought on board.” Compliance, ethics and shared strategies also play an important role here.

Developing and utilising solutions together

Ina Laukötter, District Administrator of the Gütersloh district, also emphasises collaboration. Together with the University of Bremen, the district has drawn up its own AI strategy and usage guidelines. “We need to get as many staff members on board as possible, through training, but also through role models at the top. And we need to share solutions across local authority boundaries,” says Laukötter.

This also involves moving away from the search for the ‘perfect’ individual solution for one’s own organisation. “Instead of continuing to paint lily of the valley in chequered patterns, we should rather join forces with others and use standardised solutions across the board.”

Jörg Krämer, Deputy CIO in the Administrative Digitalisation Department at the Hessian Ministry for Digitalisation and Innovation, sees AI as a historic opportunity: “When it comes to other issues, different administrations each have their own history. But with AI, everyone is starting from scratch.”

In addition to the right tools, building expertise and establishing governance are crucial. “It’s about developing ethical principles and regulations, as well as guidelines for staff. But it’s also important to convey a basic understanding of what AI does and how it works, thereby allaying fears and encouraging people to give it a go,” explains Krämer.

Taking fears seriously, involving managers

Research in Bremen shows that many staff members fear falling behind, being unable to keep pace with developments, or being completely replaced by AI. AI training can help alleviate these fears.

However, such training does not work for managers, explains Niehaves. “This type of training has no effect on them. Managers in other departments – such as social services or transport – fear a loss of significance, an erosion of power and losing their ability to shape decisions.”

It is therefore crucial that managers in their respective areas are able to help shape the use and limitations of AI themselves.