Token-OPEX: Inference Controls, Not the Seat Budget
Angelika Beierlein
9 Min. read time Token costs aren’t a line item in SaaS contracts. They’re variable OPEX per workflow-and ...
Read Article
7 min read
Two-thirds of German companies are working with AI, and more than half are already using it productively. But between the hype of investment and real productivity gains lies a gap that is increasingly determining the future of business locations.
Reading time: approx. 6 min.
Key Takeaways
- 🚀 Two-thirds of all German companies are actively engaged with AI, and more than half are already using it productively in real business processes.
- 🏭 SMEs are catching up: cloud-based standard solutions lower the barrier to entry, but integrating AI into existing IT architectures remains the decisive challenge.
- 🔑 To scale AI, you first need a data strategy – clean, structured, and legally compliant data is what determines success or failure.
Two-thirds of all German companies are actively engaged with artificial intelligence, and that share is growing. What was long seen as a future topic is now operational business. The latest Bitkom survey of more than 600 companies with 20 or more employees paints a clear picture: AI is the single strongest driver of digitalization in the German economy, across all sectors.
This is not a given. For years, Germany lagged behind in digitalization. Now a course correction appears to be underway, driven not by regulatory pressure but by companies’ own economic self-interest.
The Digitalization Index is rising again
After a period of stagnation, the Bitkom Digitalization Index is climbing once more. The increase can be directly linked to the growing integration of AI in businesses. More than half of the companies surveyed are already using AI in production, no longer just in pilot mode but in real business processes.
KPI 1
41 %
are using AI.
KPI 2
52 %
report benefits.
KPI 3
66 %
want to expand use.
“41 % of German companies are using AI, over half report competitive advantages, and two-thirds plan to expand their use. This shifts the playing field for IT decision-makers fundamentally.”
This shifts the playing field for IT decision-makers fundamentally. Those who have been waiting are now under pressure. Early adopters report measurable efficiency gains across the board. As the pace of AI adoption in the German economy shows, the leap from pilot to production is the decisive step.
Where AI is being used in practice
Adoption is concentrated in three core areas: automating routine tasks, enhancing data analytics, and supporting customer service. The difference from earlier IT projects lies in the speed and scale of implementation.
Companies that once needed weeks to analyze market data now complete the task in hours, thanks to AI-powered analytics tools that no longer require data scientists. The democratization of data analysis is one of the most visible AI impacts at the enterprise level.
AI adoption is especially strong in communications and marketing. Text generation, automated customer interactions, and intelligent search functions are low-threshold applications that deliver measurable results quickly. No wonder adoption rates are highest in these areas.
Be honest: how many of your departments are already using AI tools – with or without your knowledge?
Large enterprises lead, but SMEs are catching up
Bitkom’s data reveals a familiar pattern: large enterprises continue to pull ahead. They have dedicated AI teams, higher investment budgets, and robust data foundations. Yet the gap with SMEs is narrowing. Increasing numbers of mid-sized companies report running AI applications in production, often via cloud-based standard solutions without their own infrastructure.
This is strategically significant. For IT leaders in SMEs tasked with justifying why AI investments are urgent, these figures provide solid backing. Citing competition is no longer a scare tactic – it’s a verifiable market response.
The sticking point is integration. Deploying AI tools in isolation yields little value. They must be embedded into existing landscapes – ERP, CRM, data storage. Many projects fail here not because of the technology, but because of IT architecture.
Skills shortages persist, but less than feared
One surprising survey finding: while skills shortages are still cited as a barrier, they’re losing ground as the primary brake on digital transformation. AI tools are partially offsetting missing capacity, notably in software development, IT support, and document processing.
This shifts the conversation. The question is no longer just “How do we find qualified staff?” but “How do we enable our existing staff to use AI effectively?” Those who grasp this early build an advantage that’s hard to replicate. The companion piece on 149,000 unfilled IT roles and how CIOs are using AI copilots as skill substitutes shows how far this trend has already spread.
In practice, adoption is uneven: many employees greet AI tools with skepticism, driven by fear of job loss or sheer habit. Change management therefore becomes a technical core task, not an optional add-on.
Trust and data protection remain critical variables
Beyond technical integration, surveyed companies cite trust and data protection as key hurdles. It’s not just about GDPR compliance in the narrow sense. It’s about which data firms feed into external AI models and what control mechanisms they put in place to govern that flow.
To use AI productively, you need a data strategy – not an AI strategy. Access to clean, structured, legally sound data determines the success or failure of AI projects more than model or vendor choice.
Firms investing in data quality, data governance, and clear ownership models are laying the real groundwork for AI at scale. Everyone else buys expensive tools and wonders why results disappoint. For deeper guidance, see the article on data-driven decisions translated into C-level strategy for concrete frameworks.
What IT leaders should do right now
The Bitkom figures are piling on the pressure to act. Three common mistakes: launching AI projects without clear business objectives, failing to transition pilots into production, and not auditing the IT architecture before rolling out AI.
The next step isn’t another pilot – it’s a structured AI rollout with measurable KPIs. If you’re still in the evaluation phase, ask yourself whether your strategic rationale still holds or if it’s simply hesitation.
German companies have grasped that AI isn’t a standalone tech project – it’s the foundation on which future digitalization will be built. Which businesses benefit won’t be decided in the AI lab, but in the CIO’s decision on when to stop observing and start acting.
Frequently Asked Questions
▸What does the latest Bitkom survey reveal about AI adoption in Germany?
The Bitkom survey of over 600 companies with 20+ employees shows that two-thirds are actively engaged with AI, and more than half are already using AI productively in real business processes. The Digitalization Index has risen again after a period of stagnation.
▸In which areas are companies most frequently deploying AI?
The top three use cases are automating routine tasks, enhancing data analytics, and supporting customer service. Text generation and automated customer communication in marketing show particularly high adoption rates.
▸Can mid-sized companies benefit from AI without their own infrastructure?
Yes. More mid-sized companies are deploying AI productively, often via cloud-based standard solutions without building their own infrastructure. The gap with large enterprises is shrinking, but integrating AI into existing IT architectures (ERP, CRM) remains the key challenge.
▸Why is the skills shortage becoming less of a barrier to AI adoption?
AI tools are helping fill gaps in areas like software development, IT support, and document processing. Meanwhile, companies are shifting focus: instead of hiring new talent, they’re upskilling existing staff to work with AI tools.
▸How critical are data privacy and trust in AI adoption?
According to Bitkom, data privacy and trust are among the biggest hurdles. It’s not just about GDPR compliance – companies also need to control which data flows into external AI models and establish robust governance mechanisms.
▸Why is a data strategy more important than an AI strategy for scaling AI?
Clean, structured, and legally compliant data is the make-or-break factor for AI success – far more than model or vendor choice. Companies lacking solid data governance often see disappointing results even with expensive AI tools.
▸What are the most common mistakes in enterprise AI rollouts?
The top three pitfalls: launching AI projects without clear business goals, failing to scale pilots into production, and not auditing IT architecture before rollout. Today’s best practice is a structured AI deployment with measurable KPIs – replacing the traditional pilot approach.
Further Reading
Image source: AI-generated (June 2026)