Closing the AI Skills Gap in Companies | Alphabees

The TÜV Training Study 2026 paints an ambivalent picture of the German economy: On one hand, 27 percent of companies have already trained their employees in using artificial intelligence – double the figure from 2024. On the other hand, 56 percent of companies are already using generative AI applications in their daily work. Training is not keeping pace with adoption. For education providers, academies, and training managers, this presents a strategic challenge: How can this skills gap be closed with available resources?

The Gap Between Adoption and Competence Is Growing

The study results highlight a structural problem: While AI tools like ChatGPT, Gemini, or Copilot have become part of daily work in many companies, systematic training lags behind. 45 percent of surveyed companies currently see no need for AI training. This assessment contradicts actual usage patterns and carries risks: Employees are using tools whose functionality, limitations, and implications they only superficially understand.

The discrepancy is particularly evident across company sizes. While nearly half of large companies with 250+ employees have already conducted AI training, the share among small companies with 20 to 49 employees is only 21 percent. The Mittelstand – traditionally the backbone of the German economy – risks falling behind in digital transformation.

Training Important, But Rarely Strategically Anchored

The study also shows that training remains highly important for 87 percent of companies. Yet only 29 percent have a written training strategy. This gap between perceived importance and strategic anchoring explains why many training measures occur ad hoc and are the first to be cut under economic pressure.

The framework conditions exacerbate the problem further:

• 51 percent of companies provide only three to five training days per year
• 29 percent invest less than 500 euros per employee annually
• Only 7 percent allocate more than 2,000 euros

With these resources, traditional in-person training is hardly feasible at scale for many companies. The consequence: Either only selected employees receive training, or training does not happen at all.

Scalable Learning Formats as an Answer to Limited Resources

For training providers and education managers, this situation creates a clear requirement: Training offerings must become more flexible, more personalized, and above all more scalable. The TÜV Association recommends a needs-based mix of in-person, online, and blended learning formats along with shorter learning cycles.

This recommendation addresses the core of the problem but raises a practical question: How can education providers offer individualized learning support without proportionally increasing personnel costs? Traditional e-learning formats reach their limits here, as they are scalable but often not very adaptive. Learners receive standardized content regardless of their prior knowledge or specific questions.

This is where AI-powered learning companions come into play. An AI tutor that integrates directly into existing learning management systems like Moodle can bridge exactly this gap. It is available to learners around the clock, answers individual questions about course material, and adapts its support to each learner's knowledge level. This makes personalized learning support achievable even with large participant numbers – without trainer and instructor workload growing linearly.

New Requirements for Education Providers

The TÜV study makes clear that demand for AI training will increase. 9 percent of companies are planning concrete measures, and 17 percent are currently assessing their needs. Education providers who position themselves early with scalable and quality-assured formats can benefit from this development.

This is not just about teaching AI competencies as learning content, but also about using AI as a tool in knowledge transfer itself. An AI tutor can, for example:

Support individual learning paths:

Learners receive explanations and exercises that address their specific knowledge gaps.

Increase availability:

Questions are answered immediately, even outside office hours or between in-person sessions.

Reduce trainer workload:

Recurring questions are handled automatically, allowing instructors to focus on complex consulting.

Make learning progress transparent:

Interaction data provides insights into where learners struggle and which content should be revised.

For academies, chambers of commerce, and training providers, this represents an opportunity to enhance their existing course offerings without fundamentally changing their cost structure. Integrating an AI tutor into existing Moodle courses requires no redevelopment of content but adds an intelligent support layer to existing materials.

Changing Policy Frameworks

The TÜV Association also calls for improved policy frameworks for corporate training in its study. Tax regulations, bureaucratic hurdles for using freelance trainers, and the outdated Distance Learning Protection Act from 1977 are cited as obstacles. Whether and when reforms will take effect remains to be seen.

For training managers, this means: Optimizing their own offering structure remains the most direct lever for responding to changing demand. Those who invest in flexible, AI-powered learning formats today can build competitive advantages regardless of policy developments.

The TÜV Training Study 2026 underscores a development that has been emerging for years: The pace of technological change exceeds the adaptability of traditional training structures. The AI skills gap is just one symptom of a more fundamental problem. Education providers who design their formats to be scalable and adaptive can not only address this specific gap but position themselves for the future overall. Using AI tutors in learning support is not an end in itself but a pragmatic response to real resource constraints in the education market.

Frequently Asked Questions