Copyright OwnershIP In Tanzanian Collaborative Filmmaking Projects Using Deepfake Technologies.

06 Mar 2026 --
0 Comments

📌 I. Tanzanian Copyright Law: Framework and Challenges with Deepfake Films

1. Tanzanian Copyright and Neighbouring Rights Act, 1999

Tanzania’s core copyright statute is the Copyright and Neighbouring Rights Act, 1999. It protects “original works” including **cinematographic works and audio‑visual works”, stating the author has exclusive rights by virtue of creation. Importantly, the Act only recognizes a natural person as an author and grants moral and economic rights exclusively to them.

➡️ Key point: Under the 1999 Act, AI cannot be an author and thus works autonomously generated by AI (including deepfake films) fall into a legal gap — no clear statutory owner exists under the law. The statute does apply to audio‑visual content, meaning human‑directed film projects still fall under its scope, but it does not contemplate AI as a creator.

📌 II. International and Comparative Case Law on Deepfakes, Copyright and AI

Since Tanzania has no reported deepfake case law, we turn to relevant international litigation and judicial decisions demonstrating how courts have dealt with similar issues of copyright, authorship, personality rights, and deepfakes worldwide.

🧑‍⚖️ Case 1 — Moscow Arbitration Court: Deepfake Video As Subject of Copyright

A40‑200471/23 (Russia, 2023)
A corporate dispute involved a video created via deepfake technology that used an existing actor’s image. The court **recognized the deepfake video as a copyrightable work because:

A script, direction, audio composition and creative choices were made by humans.

AI was merely a tool, not the author.

The deepfake output was a derivative of pre‑existing copyrighted content.**

The court held such deepfake works are protected and cannot be used without consent from the original rights holders.

Takeaways for Tanzania:

Human creative input remains essential for copyright, even in AI‑enabled works.

Deepfakes don’t create a new copyright owner unless creativity and direction are human.

🧑‍⚖️ Case 2 — Garcia v. Google, Inc. (USA, 2013)

Garcia v. Google, Inc., C 12‑61628 WHA (US District Court)
Actress Cindy Lee Garcia sued Google alleging that a film included her performance (deepfake‑style reuses) without consent. The US court examined whether an actor’s performance itself is copyrightable:

Held that her individual performance can be copyrightable if original and fixed.

The court looked at human creative elements (expressive contributions) as protectable.

Relevance to deepfakes:
If AI generation uses a real actor's performance (voice, image), that original human contribution is protected. Deepfake reproduction without license thus may violate rights of performers.

🧑‍⚖️ Case 3 — Akira Nandan v. Sambhawaami Studios (India, 2026)

Though not Tanzanian, this recent Indian High Court ruling is very instructive for deepfake copyright and likeness issues:

The Delhi High Court barred distribution of an AI‑generated movie that replicated a public figure’s face, voice and persona without consent.

Court held unauthorized use violated privacy, personality and publicity rights — not strictly traditional copyright, but closely related to ownership of likeness.

Link to Tanzania: In Tanzania the same deepfake deployed without authorization could be actionable under privacy, defamation, or contractual rights, though copyright law lacks specific deepfake provisions.

🧑‍⚖️ Case 4 — Disney & Universal v. Midjourney (USA, 2025)

Major Hollywood studios sued Midjourney (AI tool) for generating images based on copyrighted characters. Though not specifically about films, it sets a precedent for:

How AI models trained on copyrighted content may infringe rights.

Court scrutiny focuses on the output and derivative nature of AI results, not just the training data.

This case signals that deepfake outputs that borrow copyrighted characters or scenes could face similar liability worldwide.

🧑‍⚖️ Case 5 — Naruto v. Slater (USA, 2018)

While not about deepfakes, the US 9th Circuit held an animal couldn’t own copyright. The key takeaway is:

Only natural persons can have copyright, not machines or animals.

Applied to deepfakes, this highlights that AI tools alone cannot be owners, making human direction a necessary element for copyright.

📌 III. Applying These Principles to Tanzanian Collaborative Filmmaking + Deepfake Tech

In a collaborative filmmaking project involving deepfake technology (e.g., replacing actor faces, AI‑generated voices, digital set creation), determining copyright requires analyzing:

✅ A. Who is the Author?

Under Tanzanian law, only a natural person can be an author:

✔️ The human(s) directing creative choices
✔️ Screenwriters
✔ Directors
✔ Performers whose performances are recorded
✔ Producers who coordinate creative elements

These contributors may own joint copyright if jointly creating a work for mutual purpose. The deepfake technology itself does not give someone ownership unless humans drove the creativity.

✅ B. Deepfake Components as Derivative Works

Deepfake outputs are normally derivative — they rely on pre‑existing copyrighted media (images, sound, film footage), so anyone using that without permission may infringe:

🔹 Derivative works are protected but require consent from original owners.
🔹 Unauthorized deepfake generated from a copyrighted film still could infringe the producer’s rights. (Similar to Moscow Arbitration and Disney v. Midjourney reasoning.)

✅ C. Contracts and Collaborations

In collaborative projects, the first copyright owners are often defined in:

Contracts (e.g., work‑for‑hire, assignment clauses).

Agreements determining who owns what share of the copyright.

Tanzania’s Act allows contracts on future grant of rights (e.g., for commissioned deepfake work) if properly documented.

✅ D. Moral and Personality Rights

Even if deepfake film content is created collaboratively, individuals still retain:

✔️ Moral rights (right to attribution and integrity),
✔️ Privacy rights, and
✔️ Personality/publicity rights, which may be violated by unauthorized uses of likeness — these are enforceable even where copyright law is silent, as seen in Indian cases.

Tanzania does not yet have clear deepfake specific laws, but principles of defamation and personal data protection can apply.

📌 IV. Practical Example – Hypothetical Tanzanian Deepfake Film Project

Scenario:

A Tanzanian filmmaker uses AI deepfake technology to replace actor faces with historical figures in a documentary, then posts it online.

Legal Analysis:

Authorship:

AI is simply a tool.

Original Works Used:

If footage/images of historical figures are still under copyright (e.g., licensed content), using them without permission infringes originals.

Personality Rights:

Using a real person’s likeness without consent can violate privacy or reputation protections even absent deepfake laws.

Contractual Ownership:

If actors or editors sign contracts assigning rights, those dictate ownership irrespective of the technology.

No Deepfake Lead Statute in Tanzania:

Courts will likely apply general copyright, defamation, privacy and personality rights concepts, as seen in other jurisdictions.

📌 V. Summary of Key Legal Points

Issue	Likely Legal Outcome
Authorship of AI‑assisted films	Human collaborators own copyright. AI does not acquire rights.
Deepfake output using copyrighted work	Derivative work must have consent; unlicensed use is infringement.
Personality/likeness rights	Separate from copyright; unauthorized use can be restrained.
Collaborative projects	Ownership dictated by contract and extent of contribution.
Tanzanian legal gap on AI deepfakes	Law does not explicitly regulate deepfakes; existing IP/privacy law applies.

📌 Conclusion

In Tanzania, copyright in collaborative filmmaking projects involving deepfakes remains grounded in human authorship, contractual agreements, and protections afforded under existing IP and personality rights laws. Deepfakes challenge traditional frameworks because AI output alone isn’t legally recognized as an author. The key questions in disputes will always revolve around who contributed creative choices, whether pre‑existing material was used with permission, and whether personal rights were infringed. As seen from international case law, courts and legislatures worldwide are evolving their analysis to handle these issues more directly.