The digital association Bitkom expects AI spending by German companies to reach 28.7 billion euros by 2026 – an increase of 48 percent compared to the previous year. The figures published on Tuesday come from an evaluation by market researcher IDC. For 2027, the association anticipates another increase of 40 percent to 40.3 billion euros.
Generative AI doubles its spending
The biggest growth driver is generative AI: Applications like ChatGPT, Microsoft Copilot, or Google Gemini will cost German companies around 11.5 billion euros in 2026 according to Bitkom – double the amount from the previous year at 5.7 billion euros. This means that generative AI now accounts for about 40 percent of all AI spending in the country, compared to just under 30 percent in 2025. The jump indicates a shift from test projects to paid widespread use: Many companies initially tested generative tools through individual pilot teams or private access for employees.
In 2026, the budget will increasingly flow into centrally managed corporate licenses with access control, data protection requirements, and integration with internal systems such as email, spreadsheets, or customer databases. For employees in offices, this primarily means: More employers are providing official AI assistants for text, spreadsheets, and communication, instead of referring employees to their own test accounts. This also changes how departments plan for AI costs – no longer as an experiment with leftover budget, but as a fixed, recurring item in the IT annual budget.
Software remains the largest expense item
Broken down by market segments, 16.0 billion euros will go to AI software, an increase of 65 percent compared to the previous year – by far the largest and fastest-growing item. Services related to AI projects, such as consulting, integration of existing systems, and training of personnel, amount to 6.8 billion euros, an increase of 30 percent. Hardware such as servers, storage, and accelerator cards for training and operating models accounts for 5.8 billion euros, a growth of 32 percent.
The fact that software is growing faster than services and hardware combined indicates a maturing market: Companies are increasingly purchasing ready-made products instead of buying expensive project work or building their own infrastructure. IDC collects the figures as part of the global AI and Generative AI Spending Guide and represents Germany as its own country segment, which is regularly updated. The specific euro amounts are based on a modeling by the commissioned market researcher based on surveys and market data and are independently unverified.
Bitkom sees transition from testing phase to practice
Bitkom President Ralf Wintergerst interprets the figures as a sign of maturation: In AI, one is experiencing “the transition from the experimental phase to widespread use” in companies of all sizes and in all industries. The association had just reported a few days ago that 57 percent of German companies are now using AI, three times as many as two years ago. The new figures provide the financial counterpart: More usage also means higher ongoing costs for licenses, infrastructure, and consulting.
The data was published on the occasion of the AI, Data and Quantum Summit, which takes place on Tuesday and Wednesday in Berlin, bringing together politics, associations, and companies on AI topics. Meanwhile, the stock markets view the growth dynamics more skeptically than in the Bitkom statement: Rising financing costs and cheaper Chinese models are fueling concerns about overcapacity in data centers – a contrast to the optimistic spending forecasts for the German market.
It will be crucial whether the projected 40.3 billion euros for 2027 will actually be reached or whether the concerns about overcapacity discussed in the capital markets will slow down the spending dynamics. A reliable interim assessment is unlikely to be provided until Bitkom’s next survey in early 2027.
