Copyright Compliance

Copyright must always be considered (see here (external link, opens in a new window)).

Accordingly, when using pre-trained AI models, care must be taken to avoid copyright infringement both when entering content (text, images, videos, audio files, code, etc.) within the prompt or other interfaces and when using results. It should also be noted that AI-generated works do not enjoy copyright protection.

For the automated analysis of content, e.g. in the context of text and data mining by the AI, it should be noted (following from §44 and §60 of the Copyright Act) that the reproduction of content, e.g. by copying in local data storage, is only permitted if the works are legally accessible and the rights holder has not excluded this form of use.

For research purposes, the reproduction of legally accessible works is permitted within this framework, provided that no commercial purposes are pursued. The processed data must be deleted as soon as it is no longer required to achieve the research objective.

This and further information on compliance with copyright law and the protection of confidentiality can also be found in the guidelines for authorities (external link, opens in a new window) and employees (external link, opens in a new window) of the Bavarian State Government and the AI-guideline for higher education (external link, opens in a new window) of the university of applied sciences Bavaria.