Copyrighting your face or voice as a deterrent to deepfakes, while conceptually compelling, raises serious legal, ethical, and practical questions particularly in the context of Trinidad and Tobago, where such protections are not yet explicitly recognized in the current intellectual property framework.
As explored in the Fast Company article, Denmark is testing the idea of letting individuals copyright their likeness and voice, not just their creative work. The rationale is that deepfake content violates personal ownership of one’s biometric identity and if such identity traits are copyrightable, creators of unauthorized AI content could be sued or blocked more easily. But can this work here in T&T?
Trinidad and Tobago’s Copyright Act (amended 2008) largely protects original literary, artistic, and musical works. It does not currently include biometric identity or likeness like your face or voice under copyrightable subject matter. The law does not prevent the creation or manipulation of someone’s image or voice unless it uses a protected work (like a copyrighted video or song).
Celebrities, athletes, musicians and other public figures may have commercial rights of publicity or legal recourse under defamation, passing off, or privacy laws, but these are not explicitly listed under copyright.
The average citizens, including minors and vulnerable persons, have even fewer avenues of protection. A deepfake using your image in a scam or pornographic video might violate data protection or privacy principles but proving intent and harm is complex.
So is a copyright approach to deepfake a viable solution?
Enforcement is the key and copyright law is more enforceable internationally than loosely defined privacy laws. There may be a deterrent factor, knowing there are copyright restrictions could discourage some AI model trainers and content creators. Having such a legislation in theory, could protect not only celebrities but also everyday people who don’t currently have legal recourse.
However, there are challenges. Face and voice are not original works (they are facts of biology), and copyright typically protects expression not identity. Registering copyrights for individuals’ faces/voices, especially for mass public use, could be administratively burdensome and cost-prohibitive. And back to the idea of enforcement, policing unauthorized use across social media, AI models, and the dark web is already difficult even with copyright.
What we could see help us here in Trinidad and Tobago can take the form of the following:
- Amend the Data Protection Act to explicitly include biometric data and impersonation via AI-generated content, making deepfakes a privacy violation.
- Develop a distinct “Right of Persona” law similar to publicity rights in the US protecting name, image, likeness, and voice from unauthorized commercial or malicious use.
- Require disclosure labels on AI-generated content (like watermarks or metadata tagging), supported by legislation.
- Mandate AI detection tools for platforms operating in T&T to identify and flag deepfakes, especially those using real individuals.
- Public education campaigns to raise awareness among citizens about the risks and signs of deepfake scams (especially financial sextortion, cloned voice fraud, and political misinformation).
Denmark is an exception to the rule. Copyrighting your face or voice is not currently practical or enforceable in T&T, and may never be fully effective due to the nature of copyright law itself. However, the idea does signal an important shift in how we think about identity as intellectual property and it opens the door for legislative innovation.
For Trinidad and Tobago, the better path forward is to:
- Strengthen biometric privacy protections,
- Introduce anti-impersonation and AI-misuse laws, and
- Explore modernized copyright and data privacy laws that protect against malicious AI use of personal identity.
0
