📌 1. Fundamental Legal Principle: Human Authorship Is Required

Under most copyright systems today, a work must be the product of human creative effort to qualify for protection. AI systems, lacking legal personhood, are treated as tools — similar to a camera or word processor.

For example, in India, the Copyright Act, 1957 defines the “author” as a natural person who creates a work; AI alone can’t be an author. But where a human has meaningfully directed or shaped the output, it may qualify as a human‑authored work.

This principle forms the core of modern AI/copyright cases worldwide.

📜 2. Thaler v. Perlmutter (U.S.): AI‑Generated Work Denied Copyright

Jurisdiction: United States
Facts: Stephen Thaler sought copyright registration for a work made entirely by his AI system “Creativity Machine,” without meaningful human creative input.
Ruling: Both the U.S. Copyright Office and federal courts held that works created autonomously by AI are not copyrightable because copyright law requires human authorship.
Outcome: The Supreme Court declined to hear the appeal, effectively leaving the lower court’s denial intact.

Key Takeaway: If your educational VR content on Vietnamese history is produced by AI without substantial human creativity, it likely cannot receive copyright protection in jurisdictions with similar rules.

📜 3. GEMA v. OpenAI (Munich Regional Court, Germany): Training and Infringement

Jurisdiction: Germany / EU
Facts: GEMA (a music rights society) sued OpenAI, alleging that copyrighted lyrics were reproduced by the AI in responses.
Ruling: The court found that an AI can infringe copyright if it memorizes and reproduces protected works without authorization as part of its training and output generation.

Relevance for Educational VR:

If your AI model generates VR narration or imagery that reproduces copyrighted texts or visuals (e.g., exact excerpts from protected historical works), you might be liable for infringement. This ruling signals that training and output matter — not just authorship.

📜 4. Midjourney and Artists’ Lawsuits (USA): AI Training as Infringement

Jurisdiction: United States
Facts: Three well‑known visual artists (Andersen, McKernan, Ortiz) sued AI companies including Midjourney and Stability AI. They alleged that their copyrighted works were used without permission in training and that the outputed images were infringing or substantially similar.

Legal Issue: Whether using copyrighted material in training a generative AI constitutes infringement, and whether the outputs themselves infringe.

Implication: Even if VR educational content is new, the data used to train the AI influences its legitimacy — unlicensed training on copyrighted content may trigger infringement claims.

📜 5. Chinese AI Image Copyright Case (Li v. Liu – Beijing Internet Court)

Jurisdiction: China
Facts: A user created an image using AI and published it. A third party reposted it without permission, leading to a copyright dispute. Court examined whether the human user contributed enough creative input to claim authorship for the AI image.

Ruling Summary (based on commentary):

The court distinguished between pure AI output and AI work significantly modified or guided by a human.

It held that if the human provider of prompts or editing adds real creative choices, the output may be considered human authored (and thus protected).

Relevance for VR Content:
If a historian or creator actively shapes the story, dialogue, visuals, and interactions (not merely clicking “generate”), that human involvement could support copyright protection.

📌 6. Precedent from “Monkey Selfie” – Naruto v. Slater (U.S.)

Jurisdiction: U.S. Ninth Circuit (not directly AI but foundational)
Facts: A monkey took selfies with a photographer’s camera. The court held that animals cannot own copyright — reinforcing that non‑human entities are excluded from authorship.

Why It Matters:
This case has been repeatedly cited by courts when analysing AI authorship cases, because both monkeys and AI lack statutory author status.

📜 7. Indian Practice: Human‑AI Collaboration

Legal Position (India):

Under current Indian law, purely AI‑generated works are not copyrightable.

But if a human meaningfully contributes — shaping content, editing, arranging — then copyright can subsist in the parts that reflect human creativity.

Example for VR:
If a human historian or developer:

selects and sequences the historical narratives,

writes scripts for virtual interactions,

fine‑tunes AI‑generated audio and animations,

integrates outputs with human creative choices,

then that collaborative work may be eligible for copyright, with the human as the author.

🧠 8. Legal Lessons for AI‑Generated Educational VR on Vietnamese History

✅ When It May Be Copyrightable

You should design the process such that:

A human author shapes the overarching narrative and core creative choices.

AI is a tool, not the sole author.

You document editorial decisions, scripts, and creative contributions.

Then the resulting VR content could be protected, with rights vesting in the human author(s).

❌ When It Probably Isn’t Protected

If:

The AI autonomously produces VR scenes, scripts, audio, or visuals with no substantive human creative guidance.

The work is substantially identical to copyrighted source material used in AI training.
Then current law (like in the U.S., India, and much of Europe) would likely treat it as public domain or unprotectable, and you could even face infringement claims.

🧠 Practical Tips for Creators

Keep Detailed Records of Human Input

Save drafts, prompt revisions, narrative outlines, and editing logs.

These help demonstrate human creativity if challenged.

Avoid Unlicensed Source Material

Use public domain, licensed, or original content for training and integration.

Attribution and Licensing

Where AI components derive from others’ copyrighted work, secure permissions or adapt them creatively.

Consider “Fair Use”/Fair Dealing

For educational and non‑commercial uses, some jurisdictions allow limited use without permission. But this is context‑specific and risky.

📌 Conclusion

The global legal trend — especially from the U.S., European courts, and India — is clear:
➡️ Pure AI‑generated works without meaningful human creativity are not eligible for copyright.
➡️ Human guidance or modification that reflects original creative skill can tip the balance in favor of protection.
➡️ If content is too close to copyrighted sources used in training, infringement risk rises.

Applying these principles to your AI‑generated educational VR about Vietnamese history will help you plan your creative process and protect your intellectual property effectively.