On August 1, 2024, the AI regulation of the European Union came into force. Since February 2025 it has been implemented gradually and has some regulations for the use of AI systems in companies.
Responsible use of AI systems
According to the European Commission, the AI Ordinance is “focused on the potential risks of AI for health, security and fundamental rights of citizens”. It determines clear and uniform framework conditions for developers and operators of AI systems in EU countries and reduces administrative and financial expenses for companies.
Forbidden applications and competencies
The Federal Government emphasizes that “certain types of applications” that “violate fundamental rights” are prohibited. This includes, among other things, “AI systems that can be used to influence the behavior of people. Even AI-based emotion detection in the workplace, unsolicited evaluation of views of views or evaluation of social behavior are also the forbidden practices”.
It is also mandatory for affected companies in accordance with Article 4 of the Ordinance that employees have fundamental AI knowledge entrusted with AI systems. This means the ability to deal with AI systems. But it is also part of the recognition of opportunities and risks. On the basis of these skills and knowledge, employees should be able to make responsible decisions.
Practical effects on companies
Overall, the new regulation concerns all companies that participate in the development, provision or use of AI systems in the EU. In a first step, companies should check whether they are affected by the new AI regulation: providers, operators, dealers or importers of AI systems are affected regardless of the size of the company or the risk class of AI. You are obliged to further train your employees in order to meet the requirements of the regulation described above.
Particular caution applies to high-risk-KI systems because they can have significant negative consequences for the health, security and fundamental rights of people. For this reason, high-risk AI systems must be “designed and developed in such a way that they can be effectively supervised during the duration of their use of natural persons,” said Bitkom expert Holdampf-Wendel in an interview with NTV.
Incidentally, for non-affected companies, the employer has the right to issue instructions and that the use of AI tools can prohibit the workplace. Use employees of AI tools without the consent of the employer, this is a violation and can be rated as a breach of duty: “The consequences range from admonition to a warning to termination,” emphasized Prof. Michael Fuhlrott, specialist lawyer for labor law and member of the Association of German Labor Law Attorns (VDAA), to NTV.
Editor finance.net