The same technology that makes a digital spokesperson possible — face transformation, voice cloning, synthetic speech — is also the technology behind deceptive deepfakes. The line between the two isn’t the technology itself. It’s consent, disclosure, and how the tool is actually used. Getting this right isn’t just a compliance exercise; it’s what protects the trust your spokesperson is supposed to be building in the first place.
This page is general information, not legal advice. Requirements vary by jurisdiction and change over time — consult a lawyer familiar with your specific market before finalizing your approach.
Consent Is the Foundation
Using someone’s likeness or voice — whether that’s a hired presenter, a team member, or a public figure — requires their clear, documented consent for the specific use you’re making of it. This should be true as a matter of basic fairness regardless of what any particular law requires: using a person’s face or voice in ways they didn’t agree to is the kind of thing that damages trust and reputation the moment it comes to light, for your business as much as for the person involved.
In practice, this means getting explicit written agreement covering not just the original recording, but the specific transformations you intend to apply — different languages, different scripts, different deployments — since consent for one use doesn’t automatically extend to every future use of that same footage.
Disclosure Protects Everyone, Including You
Being upfront that a video uses AI-generated transformation, voice cloning, or translation isn’t just good ethics — it’s increasingly a legal expectation in several jurisdictions, and it’s simply the more effective long-term strategy even where it isn’t required. A viewer who later discovers a testimonial they trusted was AI-modified without disclosure doesn’t just distrust that one video — they distrust the brand that used it.
This matters most for content where authenticity is the entire selling point — genuine customer testimonials, in particular. If a testimonial is transformed or translated using this technology, disclosing that clearly is what keeps the format trustworthy rather than turning it into a liability the moment it’s questioned.
Where the Legal Landscape Is Heading
Regulation in this space is still evolving, and specifics vary significantly by country and by use case — commercial marketing content, political content, and impersonation of public figures are often treated very differently under emerging laws. Several jurisdictions are moving toward requiring clear disclosure of AI-generated audio or video, particularly in advertising and political contexts. Because this is changing, treat any specific legal claim (including on this page) as a starting point for a conversation with a lawyer, not a final answer.
A Practical Checklist Before Publishing Any Transformed Video
- Documented, specific consent covering the exact transformations and deployments you’re using
- Clear internal record of who consented to what, and when
- Visible disclosure where a viewer would reasonably assume they’re watching unedited content — testimonials especially
- A defined process for what happens if someone withdraws consent after a video is already deployed
- Legal review specific to each market you’re deploying into, since disclosure requirements vary by jurisdiction
Why Getting This Right Is Good Business, Not Just Good Compliance
Every argument in this guide for why this technology works — the trust a genuine performance carries, the reason a real presenter converts better than a generic avatar — depends entirely on that trust being real and not undermined the moment someone looks closer. Treating consent and disclosure as central to how you use this technology, rather than an afterthought, is what keeps every other benefit in this guide actually working.
→ Talk to our team about consent and disclosure workflows
Frequently Asked Questions
Do I need written consent, or is verbal agreement enough? Written, documented consent covering the specific transformations and uses you intend is the safer standard — verbal agreement is harder to demonstrate later if it’s ever questioned.
Is it legal to use AI voice cloning for a customer testimonial? It depends on consent and disclosure — using a cloned voice with the person’s clear agreement, and disclosing that the testimonial was AI-modified, is the responsible baseline. Consult a lawyer for your specific jurisdiction and use case.
What happens if someone withdraws consent after their video is already deployed? Have a clear internal process defined for this before it happens — generally this means being prepared to remove or replace the affected content promptly.
Does disclosure hurt conversion? Not meaningfully in most cases, and the trust cost of being caught not disclosing tends to be far higher than any small effect disclosure itself might have.