IT billion for employment agency: AI offensive and cloud expansion

The Federal Employment Agency (BA) is currently undergoing one of the most radical digital transformations within the German public administration landscape. The goal is to move away from the image of a dusty office and towards a modern, data-driven tech organization. The current budget plan for the current year 2026 demonstrates that this path is costly. IT expenditures of 992 million euros are planned. Compared to the previous year, this represents an increase of a good 100 million euros, underscoring the investment pressure on digital infrastructure.

According to the budget overview published by Table.Media, a significant portion of these additional costs will flow into the operation and flexibilization of the systems. For the item "IT rent and leases" alone, the BA has budgeted a proud 130.3 million euros for this year. This is a significant jump from the 80 million euros in 2025.

The agency cites the consistent use of SAP's "Private Cloud" offering in a German data center as the main reason for this development. The BA is thus continuing a strategy that began in 2025 with the outsourcing of central payment systems. The distribution of unemployment benefits, citizen's allowance, and child benefits now depends on a new cloud architecture designed to increase reliability and ensure scalability with fluctuating user numbers.

Drive for Efficiency and Budget Limits

The transformation is not limited to infrastructure. Particularly in the area of Artificial Intelligence (AI), it becomes clear how deeply the technology is already integrated into operational business. Currently, 32 AI-based applications are in productive use or in immediate implementation. This means that BA CEO Andrea Nahles, who has been pursuing an automation offensive for years, is only narrowly missing her self-imposed goal of 33 AI tools by the end of 2025. The fact that it ended up being one application less is not a technical failure, but a fiscal decision: the financial resources were insufficient for the 33rd project.

Ironically, this missing piece of the puzzle would have offered direct added value for many job seekers. The plan was to optimize the so-called skills catalog. This AI application was intended to elevate the precise matching in online job searches to a new level by more accurately aligning users' individual skills with employers' complex requirement profiles. However, the other 32 applications are already running in the background to free up administration from time-consuming routine tasks. Due to staff shortages, local job centers have long been relying heavily on AI.

Progress is evident, for example, at the Family Benefits Office (Familienkasse). Here, intelligent document recognition ensures that approval processes for child benefits are now partially automated. The AI can independently identify submitted enrollment certificates or training confirmations and prepare the relevant data for human case workers. Similar mechanisms are also used in processing start-up grants at the job centers. A system classifies employment contracts there and extracts the essential content. This not only optimizes internal processes but, above all, ensures faster, asynchronous feedback to customers, who then have clarity about their status sooner.

(mho)