Are Employees Allowed to Use AI Systems?
A fundamental question is whether the use of AI systems by employees to perform their work is even permissible. Under Section 613 of the German Civil Code (BGB), work performance must be carried out personally by the employee. Currently, AI systems do not (yet) perform work in place of employees. Even when employees utilize AI tools such as ChatGPT or similar systems, it remains the employee’s responsibility to perform the required work personally.
The use of AI systems is often in the employer’s interest, as they can streamline work processes and enable more efficient use of employee time. However, the authorization to use AI systems is also subject to the employer’s managerial prerogative. Before granting such authorization, the employer must carefully consider several questions, such as which systems are to be used and how to ensure compliance with applicable legal frameworks. Against this backdrop, employees should only use systems that have been approved by the employer. In case of doubt, they should consult with their supervisor.
What Should the Employer Regulate?
Employers are well advised to establish the legal framework for the use of AI systems before authorizing their deployment. Not least, the provisions issued by the (EU) legislature must be taken into account, in particular the Artificial Intelligence Act (AI Act) of the European Union. Certain provisions of this regulation have already entered into force, with others to follow gradually. For example, it is already prohibited for employers to use AI systems for emotion recognition in the workplace or in educational settings.
Additionally, employers should establish policies regarding how the use of AI systems by employees will be monitored, controlled, and reviewed. It is of fundamental importance that a responsible individual always reviews work outputs generated with the aid of AI.
Furthermore, employers must not overlook their data protection obligations. When employees use AI systems, those systems must be designed to ensure compliance with applicable data protection laws. Otherwise, there is a risk that employees, by entering personal data into AI systems, may inadvertently commit data protection violations. The same applies to the entry of trade secrets and other confidential information—there must be a clear policy outlining whether, and if so how, such data may be input into AI systems.
Employee Representative Rights
Another important issue in connection with the use of AI in the workplace involves the rights of employee representative bodies. The introduction of AI systems by the employer generally triggers co-determination rights under Section 87(1)(6) of the Works Constitution Act (BetrVG). However, this co-determination right only applies if the technical system—in this case, the AI software—is at least theoretically capable of monitoring employee behavior or performance, which is typically the case.
In addition, the planned implementation of AI systems triggers information obligations under Section 90(1)(3) of the Works Constitution Act. According to this provision, the employer must inform the works council of any planned measures concerning the use of AI systems early enough to allow the council to evaluate them and raise objections or suggestions during the planning phase.
It is also conceivable that the introduction of AI systems could trigger information and co-determination obligations under Section 111(5) of the Works Constitution Act, as this may involve the introduction of fundamentally new working methods. Whether this applies must be determined on a case-by-case basis, considering the specific system and the specific workplace in question.
Conclusion
The use of AI systems in the workplace is undoubtedly here to stay. AI systems can serve as valuable tools for relieving employees of routine tasks and making more efficient use of human resources. Work outcomes can also be improved in terms of quality. However, it is essential that employers keep the various legal and organizational challenges in mind, without which the compliant use of AI systems is not possible.
Image created by Adobe Firefly