Dated ownership record
Every registered reference work gets a content hash written to Base: cryptographic proof that this image of you existed, under your control, at a specific block height.
Creator Protection
AI likeness protection for artists and creators.
A generative model can reproduce your face from a handful of public photos, and the burden of proving what happened falls on you. Suede gives your likeness a dated ownership record and consent terms that software can read, so an unauthorized use is checkable against a public register instead of your word.
Your image is already training data.
Image models train on scraped public photos, and most creators have years of them online: press shots, album art, tour photography, social posts. Nothing about that history required your consent, and nothing in the model’s output discloses that you were in the training set. When a synthetic image of you appears in an ad, a fake endorsement, or someone else’s content, the model provider did not check with you first.
The legal right at stake is the right of publicity: your control over the commercial use of your face, image, and persona. In the United States it is state law, and it is moving. Tennessee’s ELVIS Act, signed in 2024, made it explicit that AI simulation of a person’s voice or likeness without consent violates that right. The direction of the law is clear; the practical problem is evidence. Asserting the right means proving what is yours, when you controlled it, and what you did or did not authorize.
That is an infrastructure problem before it is a legal one. A takedown request, a platform dispute, or a demand letter all start with the same question: can you show a dated record of ownership and terms? Suede exists to make the answer yes before the incident, not after.
What a registered likeness gives you.
Machine-readable consent
Your registration carries license terms that software can read: whether training is permitted, whether synthetic reproduction is permitted, and what a licensed use costs. Agents and platforms can check the terms before they act.
Provenance graph
Authorized derivatives register as children of your reference works. The relationship is public and traceable, so a licensed use is distinguishable from an unauthorized one by anyone who looks.
Licensing rail
If you choose to license your likeness, for a campaign, a virtual performance, or an AI product, the terms and the payment route are already attached to the asset. The deal executes on conditions you set.
Consent only works if machines can read it.
A statement on your website that you do not consent to AI training is a message written for humans, and no training pipeline reads it. The systems that use your image are software, so the consent layer has to be software too. When you register reference works on Suede, the license module attached to each asset states in machine-readable form what is permitted: training, synthetic reproduction, commercial use, none of the above, or specific licensed combinations with a price attached.
This matters more as AI agents start transacting on their own. An agent sourcing imagery for a campaign can query an ERC-8004 asset on Base, read the terms, and either pay the license or move on. An agent cannot do either with an unregistered photo; it can only guess. Registration is what makes your likeness legible to the systems that will otherwise treat it as free raw material.
The same terms serve you in disputes with humans. Published, timestamped consent terms remove the ambiguity that unauthorized users rely on. Nobody who checked the register can claim they had no way to know.
Frequently asked questions.
- Does registering my likeness stop someone from making a deepfake of me?
- No. No registration system can prevent a model from generating a synthetic image, and any product that claims otherwise is overselling. What registration changes is your position when it happens. You hold a timestamped, on-chain record showing which reference works are yours and when you claimed them, plus published consent terms stating what use you have and have not authorized. Enforcement still runs through platform takedowns and the legal system; the record is the evidence those processes ask for first.
- What should I register to cover my likeness?
- Register the reference works your likeness travels through: press photos, album art featuring your image, music videos, promotional footage, and any official portrait set. Each registration writes a content hash to Base tied to your account at a specific block height. Together they form a dated catalog of what your image looked like, who published it, and what terms cover it. That catalog is what you point to when a synthetic version of you circulates without consent.
- How is likeness protection different from copyright?
- Copyright covers the specific works you create, like a photograph or a recording. Your likeness itself, meaning your face, image, and persona, is covered by right of publicity, which in the United States is a matter of state law. Tennessee's ELVIS Act, passed in 2024, explicitly extended that protection to cover AI simulations of a person's voice and image. Registering works on Suede does not create new legal rights; it creates the dated ownership evidence and machine-readable consent terms that make your existing rights practical to assert.
Build your evidence first
Put your likeness on the record.
Register your reference works on Suede to get dated, on-chain proof of ownership and consent terms that platforms, agents, and courts can check.
About the author
Jason Colapietro(Johnny Suede) is the founder and CEO of Suede Labs AI. He built the creator-ownership layer for the AI media era: proof of creation, programmable IP, on-chain royalty routing, and agent-accessible licensing. Patent pending USPTO 63/947,120.