3 Questions for

Irene Bertschek

Head of the Research Department ‘Digital Economy’ at ZEW - Leibniz Institute for European Economic Research Mannheim and Professor of ‘Economics of Digitalisation’ at Justus Liebig University Giessen.

Creating value with AI: ‘Expertise and framework conditions are crucial’

In the coming years, Artificial Intelligence (AI) will change business processes in many areas and influence our everyday working lives. Its use can have a positive impact on the quality of products and services as well as on employees if automated processes create new scope for higher-value activities in everyday working life. In order for value creation through AI to be successful in its entirety, the perspective of the users must be taken into account in the implementation of AI in addition to the economic factors. Irene Bertschek explains the opportunities and challenges associated with this in an interview. She is head of the ‘Digital Economy’ research department at the ZEW - Leibniz Institute for European Economic Research in Mannheim and Professor of ‘Economics of Digitalisation’ at Justus Liebig University Giessen as well as a member of the ‘Innovation, Business Models and Processes working group of Plattform Lernende Systeme.

1

Ms Bertschek, how do you think AI can be successfully integrated into companies?

Irene Bertschek: Like other technologies, AI is not a panacea. It is therefore important that companies take a close look at which specific processes or tasks AI can help to solve problems. Several factors are crucial for AI to contribute to innovation and productivity in companies, in particular data and skills. Data should be systematically prepared and made available and be of the necessary quality to contribute to the generation of reliable results when using AI. This can relate to the use of AI for more efficient processes or recruiting as well as the integration of AI into products or services. Users should be able to assess the quality of the data and the quality of the results of an AI application in order to make informed decisions based on this. On the one hand, this requires specialist knowledge in order to set up and initiate corresponding processes, but on the other hand it also requires fundamental digital skills within the company. In the short term, it is therefore helpful to obtain the necessary expertise from outside, for example through collaborations. In the medium to long term, appropriate training and further education programmes should be used to build up and develop digital skills internally. Even if generative AI in particular is increasingly becoming an integral part of software applications and automatically delivers results, employees in companies should at least have a basic understanding of how AI works and be able to assess how reliable and valid the results are.  

2

What hurdles do KMUs face when using AI? How can they be overcome?

Irene Bertschek: KMUs primarily lack time and human resources when it comes to using AI. However, this is not only the case with AI, but with digital technologies in general. In the case of AI, there is also the fact that many KMUs do not yet have a concrete idea of what they can use it for. They are unsure about the expected benefits of AI and have concerns about the maturity and reliability of AI applications. Best practice examples can provide guidance both in identifying areas of application and in terms of expectations. This is because expectations of what AI can achieve are often very high. Ultimately, the use of AI is hampered by a lack of expertise within the company and a low supply of skilled workers on the labour market. Expertise in the use and handling of AI can be gained, for example, through cooperation with start-ups or scientific institutions. This can inspire concrete applications of AI and employees in KMUs can benefit from this in terms of further developing their skills. At the same time, collaborations help start-ups as providers of AI solutions to stabilise and expand their business activities. As far as the maturity and reliability of AI applications is concerned, this depends not least on the regulatory framework.  

3

What can policymakers do to support the use of AI?

Irene Bertschek: When implementing the AI Regulation adopted at EU level, it is the task of policymakers to ensure a balance between legal certainty on the one hand and the creation and utilisation of innovation potential on the other. The implementation of the provisions of the AI Regulation should be well harmonised with existing regulations, such as the DSGVO, in order to allow for consistent case law and the derivation of clear guidelines for dealing with AI, especially with regard to KMUs. The development of AI is sometimes very dynamic, as we see with ChatGPT. Regulation should therefore also remain adaptable. For example, it should allow the allocation of AI applications to certain risk classes, as provided for in the AI Regulation, to change over time. The use of real-world laboratories, which can be used to test innovative AI solutions while temporarily suspending legal rules, also contributes to the adaptability and flexibility of the regulatory framework. Finally, public administration organisations can promote the increased use of AI solutions by directly requesting them or making their data available for corresponding applications, thereby also contributing to the development of new solutions.  

Detailed expertise on the potential and challenges of using AI in companies  can be found in the white paper  ‘Creating value(s) with AI: A guide to the successful use of AI in companies’ (in German) published by Plattform Lernende Systeme.  

The interview is released for editorial use (provided the source is cited © Plattform Lernende Systeme).

Go back