Copyright Lovers

You are welcome to follow him: LinkedIn Prof. Dr. Jan Bernd Nordemann

The first German decisions are out on copyright protection of AI output. Spoiler: No surprises. Only works with sufficient human input get protection. But the burden of proof remains an interesting issue.

I presented 2 decisions of the LG (District Court) Frankfurt and the OLG (Court of Appeal) Duesseldorf to the Regional Group North of GRUR – Deutsche Vereinigung für Gewerblichen Rechtsschutz und Urheberrecht e.V. Here are my main takeaways:

• According to CJEU “Mio and Konektra”, copyright protection continues to depend on the author’s (human) personality being reflected in the work. Works created solely by AI do not enjoy independent copyright protection.
• This is confirmed for AI song texts by the LG (District Court) Frankfurt. https://lnkd.in/dHEHp3-a
• For AI made pictures, the OLG (Court of Appeal) Duesseldorf also held that they do not get protection. See AI picture of the underwater dog above.
  https://lnkd.in/d_DF8x3M
• That said, sufficient human input is possible in 2 scenarios: In the case of (human) post-adaptation of AI output. Or in the case of prompting, provided that the personality of the prompt writer is reflected in the output (see also OLG Duesseldorf).
• The burden of proof that there is sufficient human input? It is on the copyright claimant. If it is obvious that there was some AI involved – like in the underwater dog picture. As the claimant did not present the prompts, no protection for the underwater dog picture.

A big thank you to Dr. Karsten Königer for inviting me to Hamburg and to Dr. Andreas Bothe @HoganLovells for steering my talk. And for the great crowd @Bezirksgruppe Nord.