We perfect the plan and miss the moment. Why AI is now becoming the decisive competitive factor for the German Mittelstand. Guest article by Johannes Foertsch, OpenAI.
Germany is the land of engineers, precision, and tinkerers. Quality and innovation are deeply rooted in the DNA of the German Mittelstand. But the rules of the game have changed: Quality alone is no longer enough – what counts is how quickly companies bring ideas to market. Speed has become the new currency, and artificial intelligence the most powerful lever to increase it.
Studies show that the Mittelstand is starting to recognize this potential. According to KfW, around 20 percent of small and medium-sized enterprises already use AI; the DIHK even puts the figure at 38 percent. Another 32 percent plan to get started in the next three years. Paradoxically, willingness to adopt AI is growing, yet the economic impact often remains low: Productivity is rising less than expected, processes remain sluggish, and many pilot projects fizzle out due to caution or uncertainty.
The cause is rarely technological – it is organizational.
AI booms privately – and fails in the office Artificial intelligence is more accessible today than ever before. Usage of large language models like ChatGPT has nearly quintupled in Germany within a single year. The younger generation in particular is driving this trend: For 18- to 24-year-olds, generative AI tools have become part of everyday life. Globally, Germany ranks among the top five countries for weekly active usage and is the European leader.
While the population increasingly takes AI for granted, many companies are falling behind. Employees use powerful AI tools privately for research, drafting texts, or creative tasks, but in the workplace they often encounter stripped-down, slow, or lower-quality solutions. The result is “Shadow AI”: Employees use AI without approval or governance because the need is real, but internal solutions are too slow. Bitkom studies confirm that AI is already being used uncontrollably in many organizations – without the necessary security standards.
This situation makes one thing clear: Introducing AI is not a classic IT project. It is a comprehensive change process that requires leadership, trust, clear governance, and targeted employee empowerment.
Why the Mittelstand is holding itself back The German Mittelstand rarely fails because of technology – it fails because of its own high standards. Data protection, security, perfection: all perfectly legitimate. But they become a brake when AI is only allowed once it is fully integrated and secured down to the last detail. While internal reviews continue, the outside world is already moving forward.
The problem is not a lack of integration. The problem is that companies never even get started. Value is not created in the concept, but in actual use. Anyone who waits until everything is perfect misses the exact moment when experience, speed, and competitive advantage are built. AI is not a project you “fully plan.” It is a capability you develop through application.
The solution is therefore not less governance, but the right governance: clear rules, secure access models, and a controlled framework in which employees can use AI immediately. Broad access beats perfect integration. Real AI literacy – and the foundation for scaling – only emerges when many people start using it. Integration remains important. But it is the second step. The first step is simply to begin.
At the same time, many companies start with small, isolated experiments: a chatbot for customer service, an automatic summary for internal memos. Yet such add-ons do not deliver a genuine efficiency leap. The big impact only occurs when entire workflows are rethought – from proposal generation to procurement and quality assurance. As long as AI is merely bolted onto existing processes, it changes little. Only when workflows are redesigned around AI’s possibilities does the desired impact emerge.
The real obstacle, however, often lies in the missing prerequisites for measurable success. Executives face pressure to deliver short-term results, while AI remains abstract for many teams. Without clear roles, fixed responsibilities, a network of champions, structured training, and transparent metrics, every project stays stuck in the experimentation phase. Even promising initiatives stall because they lack the organizational foundation for scaling and clean evaluation. This creates the paradoxical effect that companies want AI but, due to missing structures, cannot roll it out company-wide – and thus stand in their own way.
Proof instead of pilot: Where AI is already creating real value today A look at real-world practice shows how differently AI implementation can look – and still be highly effective. VfL Wolfsburg now uses more than 50 specialized Custom GPTs that automate everyday processes from marketing and HR to scouting, delivering measurable efficiency gains and six-figure cost savings. Teams create content faster, translate international communications, and summarize complex data analyses without adding resources. The Plex Coffee example shows that even small companies with distributed teams can benefit: ChatGPT centralizes all operational knowledge, answers routine questions in natural language, and cuts internal coordination effort by more than 50 percent. New employees learn via an interactive, AI-powered handbook, reducing onboarding time from weeks to just a few days. Both cases prove that value is created when AI is treated not as an experiment but as a concrete capability that visibly accelerates processes and relieves teams.
Three steps to break the AI blockade To help the Mittelstand fully leverage its innovation potential, no radical overhauls are needed – just three pragmatic steps:
- Bring AI out of the gray area and measure success through usage Instead of banning “Shadow AI,” companies should create a safe environment where employees can use modern AI tools – with clear rules, transparent data handling, and a permission system. Governance is not an obstacle here, but the prerequisite for speed. The sooner companies take this step, the faster they can manage risk and productivity simultaneously. The clear success criterion must be actual usage: a tool that employees do not use delivers no benefit to any company.
- Move from tools to workflows The Mittelstand succeeds because it masters processes. This is exactly where AI creates the greatest value: when it makes entire workflows consistently faster, more precise, and more robust. This applies to proposal and documentation processes, recruiting and HR workflows, quality and service procedures, as well as procurement and supply-chain processes. When AI is embedded directly in these workflows – instead of added as an afterthought – the productivity gains that studies often cite but rarely achieve finally materialize.
- Make employee empowerment a leadership task A common misconception is that employees need “general AI competence.” In reality, they need role-based empowerment. Effective approaches include champion models in every department, job-specific AI handbooks, standards for quality, security, and traceability, and measurement systems that make real productivity gains visible. Only when employees see how AI improves their daily work does the desired ROI emerge.
Implement instead of planning The Mittelstand does not need to reinvent AI. It needs to operationalize it. And that is precisely what it has historically excelled at: structuring processes, ensuring quality, and increasing efficiency. AI is not the enemy of these strengths – it is their natural evolution. Because Germany does not have an idea gap. Germany has an implementation gap. And AI is the strongest lever to finally close it.
About the author Johannes Foertsch is a founding member of the OpenAI DACH team based in Munich and has been Head of Mittelstand since June 2025, responsible for expanding OpenAI’s business relationships in the German-speaking region. In this role, he pursues the mission of making the potential of modern AI technologies optimally usable for companies and organizations.
Dr. Jakob Jung is Editor-in-Chief of Security Storage and Channel Germany. He has been working in IT journalism for more than 20 years. His career includes Computer Reseller News, Heise Resale, Informationweek, Techtarget (storage and data center) and ChannelBiz. He also freelances for numerous IT publications, including Computerwoche, Channelpartner, IT-Business, Storage-Insider and ZDnet. His main topics are channel, storage, security, data center, ERP and CRM.
Contact via Mail: jakob.jung@security-storage-und-channel-germany.de