Generative AI (GenAI) systems are increasingly being integrated into everyday work environments. Many organizations anticipate substantial productivity gains from these technologies. However, there is still limited understanding on how GenAI affects employees‘ work quality and efficiency. To address this gap, a joint research project was conducted in collaboration with Liebherr-Digital Development Center GmbH.

The project involved a controlled experiment in which employees performed typical knowledge work tasks, including information retrieval, content summarization, and idea generation. Participants were divided into two groups: one group was supported by a GenAI application, while a control group completed the same tasks without AI assistance. This study design allowed for a systematic comparison of quality and efficiency across conditions.

The findings show that GenAI can significantly accelerate the completion of many knowledge work tasks. However, its impact on output quality varies depending on the type of task. Notably, employees with lower initial performance levels benefit the most from GenAI support, leading to a partial reduction in the performance gap between top and bottom performers. At the same time, the results indicate that GenAI-supported work can lead to more homogeneous outputs, suggesting a potential reduction in diversity and originality.

The project provides important insights for organizations aiming to integrate GenAI into knowledge-intensive workflows. It highlights where GenAI can be most effectively applied and where risks to quality arise.