Ownership Of Synthetic Influencers Designed Using German Cultural Datasets.
📘 I. Legal Framework: Synthetic Influencers in Germany
Synthetic influencers are AI-generated digital personas used for marketing, media, or social purposes. Ownership and protection in Germany intersect copyright law, personality rights, and database rights.
1️⃣ AI-Generated Works and Human Authorship (§ 7 UrhG)
Only humans can be authors.
Purely AI-generated images, videos, or text without significant human creative input are not protected under German copyright.
Implication:
Whoever programs, curates, or designs the synthetic influencer with substantial human input may claim authorship.
2️⃣ Personality Rights (Recht am eigenen Bild / § 22 KunstUrhG)
Synthetic influencers designed with identifiable human traits may invoke personality rights if modeled after real individuals.
Permission is required to use likenesses.
Violation can lead to claims for injunctions and damages.
3️⃣ Database Rights (§ 87b UrhG / EU Directive 96/9/EC)
German law protects structured datasets (cultural datasets, images, music).
Unauthorized use of copyrighted datasets for training AI may constitute infringement, even if outputs are new.
4️⃣ Trade Secrets and AI Models
Proprietary AI models and training datasets may be protected under trade secret law if confidential.
⚖️ II. Key German Case Law Relevant to AI-Generated Influencers
1️⃣ GEMA v. OpenAI – Munich Regional Court, 2025
Facts:
AI-generated outputs reproduced copyrighted music and lyrics.
Holding:
AI outputs reproducing copyrighted material without license constitute infringement.
Liability may fall on the AI provider.
Implication for Synthetic Influencers:
Using copyrighted German cultural datasets (e.g., music, images, literary text) to design AI personas requires authorization.
Ownership of outputs may be contested if derived from protected works.
2️⃣ LAION / Robert Kneschke – Hamburg Regional Court, 2024
Facts:
Training datasets included copyrighted images scraped from the web.
Holding:
Training for AI falls under Text and Data Mining (TDM) exceptions if for research.
Publicly available images used in AI datasets may be lawful if they comply with machine-readable opt-outs.
Implication:
Synthetic influencers trained on copyrighted visual datasets require careful licensing unless used for educational or research purposes.
3️⃣ Metall auf Metall – Federal Court of Justice (BGH), 2016–2019
Facts:
Music sampling case where even small portions reproduced protected work.
Holding:
Recognizable portions of copyrighted works are protected.
Implication:
AI influencers incorporating distinctive cultural elements (songs, artistic motifs, visual styles) may inherit copyright risk.
Ownership of outputs is limited if they reproduce protected elements.
4️⃣ Langenscheidt / Brockhaus – BGH, 1990
Facts:
Condensed dictionaries claimed to infringe copyright.
Holding:
Summaries reproducing the essence of the original work are derivative works and require permission.
Implication:
Synthetic influencer content (e.g., AI-written bios, posts) based on cultural datasets may be derivative works, limiting ownership rights.
5️⃣ VG Wort v. Spiegel Online – Cologne RC, 2018
Facts:
Automated summarization of news articles published online.
Holding:
Even paraphrased outputs reproducing core expressive content infringe copyright.
Implication:
AI-generated dialogue, narratives, or posts for synthetic influencers must avoid reproducing distinctive texts from cultural sources.
6️⃣ Alcolix Case – BGH, 1994
Facts:
Derivative comic work claimed infringement.
Holding:
Free use (freie Benutzung) differs from derivative work (Bearbeitung).
Adaptations require permission unless transformative.
Implication:
Designing AI influencers using German cultural motifs (visual, musical, literary) may be permissible only if outputs are sufficiently transformative.
7️⃣ Mephisto Novel Adaptation Case – BVerfG, 1971
Facts:
Novel adapted from real-life events, balancing freedom of expression with personal rights.
Holding:
Artistic freedom is protected but must respect personal rights.
Implication:
AI-generated influencers modeled after real individuals must respect personality rights and privacy.
Ownership may be limited if rights are violated.
📑 III. Ownership Analysis for Synthetic Influencers
| Factor | German Legal Implication |
|---|---|
| Human input in creation | Only humans contributing substantial creative choices (visual design, textual content, voice modulation) can claim authorship. |
| Training dataset copyright | Use of copyrighted German cultural datasets may restrict ownership; license required. |
| Derivative content | Outputs reproducing protected elements are derivative; original authors retain rights. |
| Personality rights | Using traits of real individuals limits commercial exploitation; consent is required. |
| Public domain content | Works older than 70 years can be used freely for AI influencer creation. |
| Transformative use | Sufficiently transformative AI-generated content may allow human creators to claim ownership. |
⚖️ IV. Practical Guidelines for Ownership of AI Influencers in Germany
Document human contribution
Record design, text selection, stylistic choices, and narrative curation.
Use licensed datasets
Ensure that all German cultural datasets (images, text, music) are cleared for commercial use.
Avoid identifiable real individuals
Respect personality rights (§ 22 KunstUrhG).
Focus on public domain or transformative content
Goethe, Schiller, classical music, and historical imagery are safe.
Claim ownership carefully
Ownership lies with humans who materially contribute to creative design; AI alone cannot hold rights.
🏁 V. Conclusion
Synthetic influencers are at the intersection of copyright, AI law, and personality rights in Germany.
Ownership is limited to human creators who contribute substantial creative input.
Using copyrighted German cultural datasets without authorization can prevent ownership claims.
Transformative and public domain content increases the likelihood of lawful ownership.
German case law consistently emphasizes recognizability, derivative work, and human authorship as central factors in ownership disputes.
RELATED Blog
comments