Next-Generation Public Administration: Why we need the digital state as a platform

For more than 20 years, one digitalisation programme has followed another in Germany. Portals, online forms and digital administrative services have long been in place – yet citizens and businesses are still unable to interact with the state entirely digitally. The main reasons for this are a shortage of skilled workers, federal fragmentation and outdated technical infrastructure.

At the same time, generative AI and AI engineering are opening up new opportunities to accelerate development and migration. So should the government continue to extend existing specialist procedures – or is now the time for a targeted fresh start?

Our proposal: Germany should consistently view the digital state as a platform. 

The Online Access Act (OZG) provided an important impetus: administrative services should be made available online, in a user-friendly manner, via digital portals. However, additional online services alone have not solved the structural problems.

Many digital services today operate alongside one another and, in some cases, independently of one another. Different portals, specialist procedures and areas of responsibility need to be linked together. Behind modern user interfaces, old, monolithic systems and manual processes sometimes still underpin operations. The digital façade is new, but the core processes behind it often are not.

Nor was the ‘One for All’ (EfA) principle able to realise its full potential. The idea that a solution developed by one organisation can be adopted by others makes sense; however, in practice, there was often a lack of common architectural specifications, technical compatibility, integration capabilities and reliable funding.

The result: rather than genuine scalability, this continues to lead to duplicate development, siloed solutions and additional integration work.

The conclusion to be drawn from this is not that digitalisation does not work. Rather, it is that the underlying model needs to change.

The key shift in perspective lies in no longer treating individual administrative services or portals as the starting point for digitalisation. Instead, a shared platform should form the technical and procedural foundation.

A portal shows where a service can be found and what information is required for it. A platform ensures that it can be implemented, integrated, reused and managed throughout the entire process.

In practical terms, this means that users no longer have to worry about the question ‘Which form do I need?’, but can focus on their specific enquiry:

• I've moved house.
• I’ve had a baby.
• I would like to register a business.
• I’d like to build something.

The platform translates this requirement into the necessary services, data, supporting documents, responsibilities and process steps. In the proposed vision, there will then be a single digital point of access for citizens and businesses, supplemented by accessible web, telephone and public authority channels. From searching, to filling in forms, to submitting them – everything is handled via a single platform.

Underpinning the system is a shared platform core – a sovereign eGov hub. It provides functions that currently have to be developed or integrated from scratch time and again in many processes: identity and authentication, email accounts, payments, retrieval of evidence, connection to registers, status tracking, process control, monitoring and interface management.

Technical responsibility remains at federal level and specific to each sector. The federal states and local authorities do not relinquish their powers. However, their specialist procedures and services are placed on a common technical footing.

A platform of this kind can only work if interoperability does not have to be added as an afterthought, but is built into the architecture from the outset.

Standards such as FIM and XÖV provide the technical and semantic foundation. OpenAPI and AsyncAPI enable standardised technical interfaces and event-driven communication. Identity and trust standards such as eIDAS, EUDI-Wallet, OAuth and OpenID Connect provide the necessary layer of trust. NOOTS (National Once-Only Technical System) supports the digital retrieval of evidence from registers.

This gives rise to an important principle: new administrative services are no longer developed as stand-alone projects, but as building blocks of a shared digital ecosystem.

This also leads to a rethink of reuse. Instead of transferring entire individual projects from one country to another, a shared marketplace could provide reusable services, interfaces, process models, agents, form schemas, adaptors and other components. In this way, ‘One for All’ transforms from a political promise into a technical platform function.

Perhaps the biggest change compared with previous approaches to digitalisation lies in the role of AI. AI can help citizens to formulate their enquiries, find the services they need, understand forms and provide supporting documents. On the administrative side, AI can structure processes, analyse documents, verify supporting documents, initiate specialist procedures and prepare decisions.

AI thus becomes an assistance and orchestration layer within the public administration. One possible scenario, for example, involves the interaction of various specialised agents: an intent agent identifies the enquiry, a service agent assigns it to a specific administrative service, a form agent assists with the application process, and a documentation agent determines which supporting documents are required. Other agents can coordinate access to registers or integrate existing specialist procedures.

Governance is key here: AI may prepare, structure, review and prioritise processes. However, the final legal decision remains with the relevant human. Model gateways, authorisations, logging, audit trails and escalation rules must ensure that its use is traceable and controllable.

One of the biggest challenges remains the legacy technical infrastructure. However, these systems should not be taken offline all at once, but rather migrated systematically as part of a migration strategy. This could be carried out as follows: The platform proposed here first connects new and existing systems via a standardised API layer. This allows business processes to be classified step by step:

• Integration: The existing process will remain in place for the time being and will be integrated via standardised interfaces.
• Refresh: A process that is strategically important but technically outdated is redeveloped using modern architecture.
• Consolidation: Several similar processes are combined into a single service.
• Decommissioning: Systems that are rarely used or for which an alternative platform is available will be taken offline.

Such a transformation must be carefully planned, and a platform strategy does not mean replacing the existing administrative IT systems all at once. The transition can take place in stages.

1. Phase: Foundations and pilot projects: Initially, the platform core, eGov hub, API gateway, identity and authentication layer, and service catalogue will be developed. In parallel, three to five administrative services will be implemented as pilot projects, such as change of address, driving licences, business registration, parental allowance or planning permission.
    
2. Phase: Scaling and Migration: Successful building blocks are reused; the marketplace goes live; and further specialist processes are integrated, updated or consolidated. Regions and local authorities receive defined onboarding packages, technical support and binding framework conditions.
    
3. Phase: Consolidation and stabilisation: The platform becomes the standard for new administrative services. Legacy systems that can be phased out are gradually decommissioned; the platform becomes a permanent part of the government’s digital infrastructure.

Success should not be measured by the number of new portals or individual projects, but rather by a few clear indicators: actual usage, technical integration, subsequent use and costs. 

The digitalisation of Germany’s public administration is therefore at a crossroads. It can continue along the same path as before – with more individual projects, more portals and further integration into an already highly complex system landscape – or Germany can use the changed technological landscape as an opportunity for a structural fresh start.

The concept of the digital state as a platform does not mean abolishing federalism or replacing all existing systems. On the contrary: federal responsibilities and existing specialist expertise will be retained. What is changing is the digital foundation on which they operate together.

A shared platform, open standards, reusable building blocks, AI-supported development and a consistent end-to-end approach could transform many individual digitalisation projects into a scalable administrative ecosystem.

The key question, therefore, is no longer how we digitise the next public service. Rather:How do we build the digital infrastructure on which the next 1,000 public services can be developed?