AI and Knowledge Work: Implications, Opportunities and Challenges

In the following interview, Prof. Dr Hendrik Send and Shirley Ogolla discuss the HIIG research project “AI and Knowledge Work – Implications, Opportunities and Risks”, financed by the AI Observatory of the BMAS (German Federal Ministry of Labour and Social Affairs). By studying specific applica­tions of smart and autonomous systems, they aim to identify the implications, opportunities and risks for workers and businesses.


You conduct qualitative research into the implications of AI for business practice. Why did you choose this particular subject?

Artificial intelligence (AI) technologies enable better solutions for many tasks of knowledge-based work than previous technologies have allowed. They provide radically new tools for workers for a very broad range of applications. At the same time, many people are voicing concerns about the risks of AI applications. To be able to shape technologies in a way that they meet the requirements of quality jobs and enable effective solutions in business, we need a solid understanding of the application scenarios of AI technologies.

What does your research project involve exactly?

The starting point for our project is how people who already work with AI applications perceive their work. Three topics are of particular interest to us in this regard: the intention of the companies introducing the applications, the development and implementation processes, and the individual and organisational skills associated with the application. Existing research into the use of technology in the workplace gives grounds to assume that the use of AI in knowledge-based workplaces presents an important field of research with new implications.

Which approach do you take exactly and why?

In the first phase of the project, we collect many examples of AI in practice and get a better understanding of the range of AI application scenarios that can currently be found in the German business landscape. This also involves examining which topics are raised with particular frequency by stakeholders in the context of AI. In the second phase, we conduct interviews with various experts to get a better insight into the application scenarios and the organisational context. As part of the third phase, we go to the business organisations to carry out in-depth case studies. Here, we work closely with AI users and also involve developers, project managers and other relevant stakeholders. In the final phase, we will organise pilot workshops with representatives of organisations, AI users and other disseminators to discuss the findings from our research with practitioners.

„The starting point for our project is how people who already work with AI applications perceive their work.“

Researchers expect the role of artificial intelligence in the world of work to increase significantly in the coming years. Photo: SFIO CRACHO /

What insights have you already gained?

There are two primary motivating factors behind most of the projects of our interview partners: first, many cite the overwhelming workload of knowledge-based workers as the reason for the project. Second, many explain that AI applications are part of the process of taking their business model digital and are intended to make a key contribution to their competitiveness in the future.

Most companies in our sample report that they have been pursuing a digitalisation strategy for quite some time and benefit from it when using AI. In many cases, they began recruiting specialists, such as data scientists, several years ago and have concentrated digitalisation skills in teams. In our real-life case studies, the initiative for the AI projects very often came from precisely these teams. It is remarkable how frequently these teams do not use solutions from third-party providers, choosing instead to develop their own AI applications, often using open-source software. Overall, it can be said that our companies have already deployed AI successfully and we have not heard any accounts of failed AI projects or encountered businesses with serious internal discussions or challenges.

What do you see as the most important future developments for AI in the working world and what should policy-makers, trade unions, businesses and workers bear in mind?

AI applications are already supporting, changing and replacing human work in many complex tasks. First of all, we can say that the application scenarios in our collection of cases all concern rather specific and clearly demarcated fields of activity and not combinations of multiple different tasks. In contrast, usually job profiles specifically involve combinations of different tasks. It will be important to monitor which job profiles are more affected in order to support the workers at an early stage. Second, we are seeing some AI application scenarios in which the new technology enhances the skills of the knowledge workers, and others in which sub-tasks are being automated entirely and the need for the corresponding human skills is decreasing. If possible, businesses should try to implement the use of AI in a way that the applications support human skills and contribute to further skill development. Thirdly, many of the companies surveyed in the context of AI are making explicit efforts to train existing staff in order to cover their demand for experts in the field of data science and machine learning. If businesses use the efficiency gains from AI for the benefit of everyone involved by investing the additional resources in training, then the potential for conflict associated with the further expansion of applications is minor. The high demand for experts in the field is another argument for lifelong learning.

„AI applications are already supporting, changing and replacing human work in many complex tasks.“

As scientists, we have a hard time with predictions for the future. If, however, AI applications follow the diffusion pattern of other technologies like PCs, mobile phones or the internet, after a few years of slow growth we will see an acceleration in take-up if additional innovations that can be used to complement AI applications enter the market.

share page