EU rules for artificial intelligence: competitive advantage or overregulation?

© Unsplash/Franki Chamaki

After the EU Commission presented the White Paper on Artificial Intelligence (AI) in February last year, a legislative proposal on the use of AI was published in April 2021. BDI welcomes the EU Commission's technology-open and risk-based approach. An approach based on European values can create a competitive advantage for European companies but must under no circumstances constitute an obstacle to innovation.

The EU Commission's legislative proposal assigns AI applications to different risk classes. BDI welcomes this approach in principle, but the definition of so-called high-risk applications in the Commission's draft regulation is too broad. As a result, less risky industrial applications would also fall within the scope of the comprehensive requirements for high-risk AI.

The definition of high-risk applications is too broad

This would be disproportionate and would deter small and medium-sized enterprises (SMEs) and particularly start-ups from developing innovative industrial AI applications. The measures planned by the Commission to promote innovation in the field of AI are fundamentally positive. These include, for example, enabling so-called regulatory sandboxes. However, the measures presented to promote this do not go far enough and should be expanded in order to ignite the innovative power that AI offers in all companies. After all, innovation-friendly rules in the EU help European companies develop AI-based business models and survive in the global innovation competition. At the same time, they favor the establishment of European standards on a global level.

A balanced legal framework can become a competitive advantage

Eventually, the further legislative process must therefore be aimed at finding the right balance between requirements for undoubtedly risky AI applications and innovation-friendly and unbureaucratic framework conditions for companies. In doing so, it is essential that the legislator also considers the industrial scope of application of AI. It is also important for the further promotion of the use of AI that comprehensive information is provided about the use of AI. In many companies, AI is still thought of in a too complex way. Yet there are numerous examples of possible applications with low complexity: AI can be employed for automatic error detection in production or it can be used in customer service to classify e-mails or in the context of chatbots to answer standard customer inquiries quickly. As a technology of the future, AI can make an important contribution to solutions in a wide variety of areas of society, such as achieving sustainability goals. That is why the innovative power of this key technology, with all its positive effects on life, society and climate in Europe, should be fully exploited.

In the further legislative process, BDI will advocate ensuring that the new EU rules for AI are designed in an innovation-friendly way by the European Parliament and the member states.