 Artificial intelligence (AI) has made its way into the everyday work of many organizations. Hardly any other technological development in recent decades has been adopted so quickly and at the same time been the subject of such controversial debate. While proponents expect enormous productivity gains and new forms of value creation, critics warn of job losses, de-skilling, and increasing stress for employees. As is so often the case, the reality lies somewhere in between. In this article, I classify current empirical findings on the performance and use of AI, analyze its effects on individuals, organizations, and the labor market, and derive implications for companies and employees. 1. AI is rapidly becoming more powerful
The performance of modern AI systems has increased at a remarkable pace in recent years. Current analyses show that the duration of tasks that AI agents can perform autonomously with a 50% probability of success has doubled every seven months for about six years. This means that the limits of what can be automated are constantly shifting upward, especially for knowledge-intensive activities. AI systems already outperform humans in many text, analysis, and knowledge tasks, often even experts – and at a fraction of the cost. This development largely explains why AI is no longer perceived solely as an assistive technology, but increasingly as an independent “player” in work processes. 2. AI use in companies is increasing significantly Parallel to its increasing performance, the use of AI in companies is also growing. Current Eurostat data shows that almost all EU member states have recorded an increase in the proportion of companies using AI compared to the previous year. This growth is particularly strong in Northern Europe, for example in Denmark, Finland, and Lithuania. AI is currently most commonly used to analyze written language, followed by the generation of text, images, audio, or video content. These usage patterns illustrate that generative AI is no longer a niche topic, but is permeating central office, communication, and knowledge processes. 3. From “nice-to-have” to strategic partner in everyday work With increasing availability, the nature of AI use has also changed qualitatively. Studies show that daily use of AI tools has nearly doubled within a year. At the same time, the proportion of those who consider AI useless has fallen dramatically. More and more employees no longer view AI as just a tool, but as a strategic cooperation partner. This development can be interpreted as a maturation process: from occasional use to simple assistance functions to collaborative integration into decision-making, analysis, and creative processes. Organizations are thus faced with the challenge of actively shaping this new human-machine interaction instead of leaving it to individual experimentation. 4. Significant efficiency gains – but mainly at the individual level At the individual level, the effects of AI use are clearly measurable. On average, employees report productivity increases of around one-third and time savings of around 1.3 hours per working day. Managers also say that AI enables them to make more efficient decisions and improve their leadership skills. What is interesting here is that the picture is nuanced: less experienced or less qualified individuals benefit particularly strongly, as AI can partially compensate for a lack of experience and knowledge. Experienced employees also achieve productivity gains, but only if AI meaningfully complements their existing expertise rather than replacing it. In this case, AI acts as a complement, not a substitute. 5. The productivity paradox at the organizational level As clear as the effects are at the individual level, the picture at the organizational level is sobering. Despite massive investments, amounting to between $30 billion and $40 billion worldwide, up to 95% of companies report no measurable return on their GenAI initiatives. The results are highly polarized: a few organizations achieve significant effects, while the vast majority hardly benefit at all. This phenomenon is therefore also referred to as the “GenAI divide.” A key reason for this is that efficiency gains made by individual employees do not automatically translate into organizational value creation. Without adjustments to processes, roles, control mechanisms, and incentive systems, much of the potential is wasted. 6. Side effects: loss of motivation and technostress In addition to efficiency gains, there is growing evidence of undesirable side effects. Empirical studies show that AI increases productivity, but at the same time can reduce intrinsic motivation. Added to this is a phenomenon that is increasingly being discussed as “workslop”: low-quality AI-generated content that requires additional checking and correction, thus creating new inefficiencies. More and more employees are also reporting increasing stress. Although simple, repetitive tasks are being automated, the resulting free time is often filled with new demands. The result is not relief, but technostress. 7. Impact on the labor market: change rather than sudden displacement The question of whether AI will destroy jobs on a massive scale is controversial. However, empirical evidence paints a more nuanced picture. AI is replacing individual tasks rather than entire professions. Nevertheless, job profiles are undergoing profound changes as a result. Current analyses show that job advertisements for activities with a high AI impact are becoming less common. Entry-level positions are particularly affected, which is already reflected in declining employment figures for young people, for example in the US. In Germany, around a quarter of companies now expect job cuts as a result of AI. Productivity gains often lead to rationalization measures, especially in economically challenging times. AI thus acts as an amplifier of existing economic dynamics, not as their sole cause. 8. Which skills will count in the future? Against this backdrop, the question of future skills is becoming increasingly important. Technical knowledge alone will not be enough. Instead, skills that are difficult to automate will be in demand: emotional intelligence, creativity, empathy, critical thinking, and the ability to distinguish relevant from irrelevant information. Judgment, decision-making ability, and the competence to classify AI results and use them responsibly will become key differentiators in the labor market. Paradoxically, this increases the value of genuinely human skills, especially in an increasingly AI-driven world of work. 9. Conclusion: Design rather than technological optimism Artificial intelligence is profoundly changing the world of work. It increases individual productivity, is widely used, and is developing rapidly. At the same time, organizational efficiency gains remain the exception, and negative side effects cannot be overlooked. The decisive factor is therefore not whether AI is used, but how. Companies must understand AI as a socio-technical system that affects organization, culture, leadership, and skills in equal measure. Only then can the potential of AI be leveraged sustainably and responsibly. References: https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks/ https://ec.europa.eu/eurostat/web/products-eurostat-news/w/ddn-20251211-2 https://atlassianblog.wpengine.com/wp-content/uploads/2025/09/atlassian-ai-collaboration-report-2025.pdf https://www.oecd.org/en/blogs/2025/07/unlocking-productivity-with-generative-ai-evidence-from-experimental-studies.html https://www.heise.de/news/KI-steigert-Produktivitaet-aber-Unternehmen-profitieren-kaum-11067833.html https://www.artificialintelligence-news.com/wp-content/uploads/2025/08/ai_report_2025.pdf https://hbr.org/2025/05/research-gen-ai-makes-people-more-productive-and-less-motivated https://hbr.org/2025/09/ai-generated-workslop-is-destroying-productivity https://www.faz.net/pro/digitalwirtschaft/zukunft-der-arbeit/technostress-durch-ki-warum-junge-mitarbeiter-leiden-accg-200362716.html https://www.spiegel.de/wirtschaft/ki-draengt-auf-den-arbeitsmarkt-die-digitalen-agenten-kommen-a-3f7d2647-7820-4f09-96f4-d82e16daf3ad https://www.businessinsider.de/wirtschaft/ki-forscher-warnt-80-prozent-arbeitslosigkeit-ist-keine-science-fiction/ https://www.faz.net/pro/digitalwirtschaft/zukunft-der-arbeit/chatgpt-und-ki-weniger-stellenanzeigen-in-deutschland-accg-200455613.html https://digitaleconomy.stanford.edu/wp-content/uploads/2025/11/CanariesintheCoalMine_Nov25.pdf https://www.ifo.de/fakten/2025-06-05/ein-viertel-der-unternehmen-rechnet-mit-stellenabbau-durch-kuenstliche https://www.spiegel.de/wirtschaft/ki-wird-alle-jobs-veraendern-und-das-radikal-a-96c12c01-9a49-4929-aa1f-61d62d34cf6b https://www.faz.net/pro/digitalwirtschaft/zukunft-der-arbeit/wenn-unternehmen-einstellen-suchen-sie-nach-ki-kompetenz-accg-200362260.html https://www.faz.net/pro/digitalwirtschaft/zukunft-der-arbeit/chatgpt-und-ki-weniger-stellenanzeigen-in-deutschland-accg-200455613.html https://www.faz.net/pro/digitalwirtschaft/zukunft-der-arbeit/ki-kompetenzen-worauf-es-im-job-wirklich-ankommt-accg-110768933.html https://www.spiegel.de/wirtschaft/soziales/david-autor-welche-jobs-sind-auch-mit-ki-noch-sicher-a-0978af3a-3632-4ca8-b9d8-ce828f1a69de Comments are closed.
|
Archives
February 2026
|
RSS Feed