AI usage is declining: Forget the LinkedIn hype – This is how (little) Germans are really using AI

Xpert Pre-Release

Available in 27 languages 📢

Published on: August 13, 2026 / Updated on: August 13, 2026 – Author: Konrad Wolfenstein

AI usage is declining: Forget the LinkedIn hype – This is how (little) Germans are really using AI – Image: Xpert.Digital

Too expensive, too complicated? Why the AI revolution is stalling in German SMEs

The big AI check: Between digital hype and sobering reality

The AI hangover: After the initial euphoria, disillusionment is now setting in in Germany

Everyone's talking about artificial intelligence, but the reality in Germany looks surprisingly different. While social networks give the impression that AI has long since dominated our daily lives and every workplace, recent studies reveal an initial damper: The initial euphoria is giving way to a noticeable consolidation, and private usage figures are even declining in some cases. A particularly divided picture emerges within companies. Instead of relying on systematic training and clear strategies, many employees are secretly using ChatGPT and similar tools on their own – this so-called "shadow AI" is growing rapidly and becoming an uncontrollable risk for employers. This comprehensive analysis reveals why the much-discussed AI boom is currently entering a crucial phase of maturity, why Germany is nevertheless performing better than expected in international comparisons, and why millions of people simply feel overwhelmed by the technology.

Between hype and everyday life: The true use of AI in Germany: When the timeline promises more progress than reality delivers

Anyone scrolling through LinkedIn or other professional networks these days might get the impression that artificial intelligence has long since permeated every workplace, every household, and every business decision in Germany. In reality, recent representative surveys paint a much more sobering picture. The AI Germany Study 2026 by the Hamburg-based market research institute Splendid Research, for which more than 1,000 German citizens were surveyed in early summer, shows a decline in personal AI use from around 64 percent last year to 60.8 percent among 18- to 69-year-olds in 2026. According to the study, almost 40 percent of the population in this age group will completely forgo the use of AI tools. This figure contradicts the prevailing public narrative of an unbroken, exponential growth trajectory and deserves a more precise economic analysis, as it reveals an initial saturation phase within a technological cycle that has so far been portrayed almost exclusively as a story of upward mobility.

Why the perception of AI adoption differs from reality

The discrepancy between constant media presence and actual usage can be explained by classic mechanisms of technology diffusion. Early adopters, professional communicators, and knowledge workers, who are already active in digital networks, disproportionately influence public discourse, while the general population is considerably more reserved. Various studies provide sometimes significantly differing figures for overall usage, which initially appears confusing but becomes clear upon closer examination of the methodology. While the digital association Bitkom determined a usage rate of 58 percent of the population aged 16 and over for its 2026 survey, representing an increase compared to 40 percent in 2024, the Splendid Research study arrives at a declining figure of 60.8 percent for the more narrowly defined group of 18- to 69-year-olds. This apparent contradiction primarily demonstrates one thing: that snapshots from different survey periods, age groups, and questions are hardly directly comparable, and that public discourse often picks out the most spectacular figure without providing the methodological context.

A crucial factor in the differing results lies in the timing of the surveys. The Bitkom figures for 2025 come from a survey that captured the boom surrounding new language models while it was still in full swing, whereas the Splendid study from early summer 2026 may already reflect an initial phase of disillusionment, in which the initial curiosity about many AI tools had waned. Furthermore, a significant portion of the population apparently began testing AI but did not translate this into regular usage. This observation aligns with findings from the United States, where a McKinsey study identified a significant decline in weekly AI usage in the workplace from 64 percent in January 2025 to 47 percent a year later, while daily usage fell from 32 to 22 percent during the same period. The American market, long considered a pioneer in AI adoption, is thus experiencing a similar, and in some cases even more pronounced, consolidation phase.

Professional use falls significantly short of expectations

Particularly revealing is the examination of the professional application of AI tools, as this is where the real economic substance of the debate lies. The observation cited in the original article, that professional use is limited to just over a quarter of the population, is supported by several independent surveys, although the exact percentages vary depending on the methodology. A study conducted by the ifo Institute in collaboration with the Centre for European Economic Research, the Institute for Employment Research, the Federal Institute for Vocational Education and Training, and the Federal Institute for Occupational Safety and Health, concludes that only one in five employees in Germany uses AI regularly in the workplace, even though around 64 percent have already tried the tool at least once. This gap between one-time trials and regular, productive use is highly relevant from an economic perspective, because only repeated application integrated into workflows generates measurable productivity gains.

In addition, the ifo Institute study reveals a remarkable structural finding: the main application officially introduced by companies was only provided by the employer for about a third of users, while two-thirds access AI tools on their own initiative and without an official company structure. This so-called shadow AI usage is ambivalent from a business perspective. On the one hand, it demonstrates remarkable initiative on the part of the workforce; on the other hand, it creates significant risks regarding data protection, information security, and quality control, as employees often use private accounts for work purposes without the knowledge of IT departments or compliance officers. The German Trade Union Confederation (DGB) points out in a special edition of its newsletter that more than half of employees in Germany already use AI in the workplace, albeit often through unofficial channels.

Other studies arrive at differing, sometimes significantly higher, figures for professional use, further highlighting the methodological inconsistencies of the debate. A Forsa survey conducted in April 2026 by the TÜV Association found a professional usage rate of 45 percent among employed individuals, while Bitkom Research reports 48 percent for the same target group. The Boston Consulting Group even reports 75 percent regular use among employees without managerial responsibilities and 91 percent among managers, which at first glance seems hardly compatible with the ifo figures. This enormous range of results, from 20 to 91 percent, illustrates that the definition of usage, the selection of the sample, and the wording of the questions have a considerable influence on the measured result and cautions against drawing any sweeping conclusions about the maturity of AI adoption in Germany.

An international comparison puts German reluctance into perspective

Interestingly, Germany does not appear as a digital laggard in international comparison, but in some respects is even developing above average. A McKinsey study published in March 2026 found that the regular use of AI tools such as ChatGPT, Gemini, or Copilot in the workplace in Germany doubled within a year, from 19 to 38 percent, while the proportion of daily users rose from 7 to 16 percent. At the same time, the United States experienced the significant decline already described during the same period, thus at least partially diminishing the US's former pioneering role. The Boston Consulting Group concludes that Germany is now at an international level in terms of pure usage frequency, although German employees report less noticeable time savings or high levels of satisfaction compared to other countries. Only about a third of employees feel adequately prepared to use the technology, which indicates significant skills gaps.

These findings can be interpreted as an indication of a specifically German challenge: while AI adoption is certainly present in international comparison, it is insufficiently systematically supported. German companies and employees are apparently experimenting with AI tools to a similar extent as their international counterparts, but they less frequently exploit their full potential because training programs, clear usage guidelines, and systematic integration into business processes are often lacking. This gap between tool availability and skills development is likely to be more decisive for actual value creation in the long run than the sheer frequency of access to an AI model.

🎯🎯🎯 Data-driven B2B industry hub as a quasi-in-house solution

The quasi-in-house solution: How Xpert.Digital closes operational gaps in B2B marketing and sales – Smart Content-Driven Business - Image: Xpert.Digital

Xpert.Digital is a data-driven B2B industry hub led by Konrad Wolfenstein . The company acts as an external, quasi-in-house solution for industrial partners, closing operational gaps in marketing, content, and sales – without requiring additional resources on the client side.

More information here:

The quasi-in-house solution: How Xpert.Digital closes operational gaps in B2B marketing and sales – Smart Content-Driven Business

From hype to consolidation: The current state of AI usage

Companies caught between investment pressure and sobering costs

At the company level, the picture is similarly divided as with individual use. According to Bitkom, the proportion of companies using AI in their own operations more than doubled from 17 percent in 2025 to 41 percent in 2026, with 77 percent of AI-using companies reporting an improved competitive position. A study by the Institute for Employment Research confirms this rapid increase at the company level, quantifying the proportion of companies using generative AI as having jumped from 5 percent in 2023 to almost 25 percent in 2025, while a further 9 percent are planning its introduction and 67 percent currently have no plans to do so. At the same time, the same study indicates that 31 percent of employees will have used AI professionally by 2025, which the company survey considers a rather conservative lower limit of actual usage, since employees can access AI tools even without their employer's knowledge.

The rapid formal increase in the use of AI in businesses is accompanied by significant friction, which is often overlooked in public discourse. According to Bitkom, a third of companies report unexpectedly high implementation costs, while 53 percent cite legal uncertainty surrounding the European law regulating artificial intelligence as the biggest obstacle. An equal number of companies complain about a lack of expertise, and 51 percent suffer from staff shortages. Particularly noteworthy is that, according to this analysis, a traditional approach to AI implementation in SMEs can incur costs of between €250,000 and €500,000 in the first year alone, representing a considerable investment hurdle for many small and medium-sized enterprises. Furthermore, the study indicates that only 24 percent of companies have seriously addressed the implementation of the AI law, even though 69 percent state that they require support in this area. These figures paint a picture of an economy that is formally committed to AI transformation but encounters significant regulatory, financial, and personnel hurdles in its practical implementation.

Which applications are actually used

A look at specific application areas reveals that actual AI use is heavily concentrated on simple, low-threshold applications, while more complex, potentially higher-value-added applications are used significantly less frequently. The ifo study notes that individual, self-initiated use focuses almost exclusively on easily accessible text tools such as ChatGPT or automated translation programs, which are used by more than 80 percent of AI users. In contrast, formal, employer-driven use is more focused on complex and expensive applications such as diagnostic tools or speech and image processing. The TÜV Association complements this picture with the observation that around 78 percent of employed people use AI for information retrieval, 46 percent for text creation or improvement, 41 percent for idea generation, and only 19 percent for creating images or videos, while only about one in ten uses AI for programming.

This distribution reveals a fundamental economic pattern in the current diffusion of AI. The technology is primarily used as a replacement or supplement for existing, cognitively simpler routine tasks such as research and text processing, while more demanding applications, more deeply integrated into business processes, such as software development, medical diagnostics, or complex data analysis, remain the exception. Bitkom Research confirms this observation, noting that users primarily see advantages in the simplification of routine tasks, the saving of working time, and the increased freedom for more important tasks and faster problem analysis. This suggests that the macroeconomic productivity gains from AI have so far been incremental and additive rather than disruptive and transformative, which at least partially explains the cautious disillusionment expressed in current surveys.

Ambivalent feelings characterize the population's relationship with technology

Beyond the raw usage figures, another Bitkom survey provides insightful glimpses into the emotional and psychological dimensions of AI adoption, which are crucial for the economic evaluation of technology diffusion. While three-quarters of active users report enjoying working with AI, and 76 percent feel that AI makes their lives easier, 42 percent of the general population would prefer to live in a world without AI. Among those who do not currently use AI, this figure rises to 58 percent. It is also striking that around 40 percent of the general population feel left behind or overwhelmed by the technology, with these figures being significantly higher among non-users at 59 and 57 percent, respectively.

This deep-seated ambivalence explains a significant part of the observed stagnation or slight decline in usage figures. A technology that is predominantly viewed positively by active users, but simultaneously triggers feelings of being overwhelmed and excluded among a considerable portion of the population, is unlikely to achieve widespread adoption in a short time. According to another survey, 54 percent of those interviewed by Bitkom who do not currently use AI still rely on traditional internet search engines, although only 13 percent of non-users reject the technology out of deep-seated conviction. This suggests that the greatest untapped potential for growth lies not among convinced skeptics, but in a large group of undecided or simply untrained individuals who could still be won over to the technology through appropriate educational programs and low-threshold access.

What these figures mean for the German economy

From an economic perspective, the current data allows for a nuanced conclusion that avoids both pure euphoria and premature skepticism. The discrepancy between constant media coverage and actual, productive use shows that Germany is in a typical phase of technology diffusion, where an initial hype gives way to a more sober consolidation after the first wave of experimentation. This pattern is by no means unusual historically, but rather corresponds to the classic trajectory of previous technology waves such as the internet or mobile devices, where a period of inflated expectations was regularly followed by a phase of disillusionment before a sustainable, but considerably more moderate, growth curve established itself.

The future economic relevance of AI in Germany will depend on whether the currently fragmented and often self-initiated use can be transformed into systematic, company-supported applications. Data from the ifo Institute clearly shows that formal company-wide implementation is associated with higher usage frequency, more training opportunities, and greater productivity gains. This suggests that the key to unlocking economic value lies less in the mere availability of AI tools than in their structured integration, accompanied by clear guidelines, targeted training, and legal safeguards. As long as a significant number of companies continue to struggle with legal uncertainty, cost pressures, and a shortage of skilled workers during implementation, the technology's full productivity potential will hardly be realized, regardless of the sheer number of users accessing individual AI applications.

The initially provocative observation of a decline in AI usage, upon closer analysis, proves to be less evidence of the technology's failure and more a necessary and healthy maturation process. After the initial wave of playful curiosity, the wheat is increasingly being separated from the chaff, with genuine, recurring potential benefits becoming distinct from short-term curiosity. For companies, policymakers, and individual employees, this means that the real economic challenge no longer lies primarily in the technical availability of AI tools, but in the ability to filter out, from a diffuse multitude of applications, those few that actually generate measurable, lasting added value.

📈🚀 From visibility to trust 👀🤝 Your scalable path with Xpert.Digital

From visibility to trust: Your scalable path with Xpert.Digital - Image: Xpert.Digital

In industrial B2B, sustainable business relationships rarely emerge overnight. They develop step by step – through visibility, professional relevance, recurring touchpoints, and growing trust. Xpert.Digital's 4-stage model addresses precisely this: It offers a structured path that begins with a manageable entry point and can evolve into deeper collaboration in business development if needed.

Instead of relying on loud marketing promises, this model puts the relationship at the forefront. Companies start with clearly defined, easily calculable measures and then decide, based on their own experience, how far they want to expand the collaboration. A key factor for this undisturbed trust-building process: The platform completely avoids annoying advertising ads, so the editorial focus remains solely on the companies' expertise.

More information here: