Is Your AI-Cloned Voice or Face Actually Protected by UK Law?
A new offence covers non-consensual AI images since February 2026, but cloned voices and faces in fake ads still fall outside UK law. What's changing.
The short answer
Not yet, and that's the problem. Since 6 February 2026 it's a criminal offence in England and Wales to create - or even ask someone else to create - a non-consensual AI-generated intimate image of a real person. But that only covers the sexual-content end of cloning. If someone uses AI to clone your face or voice for a fake endorsement, a scam ad, or content you never agreed to, there is still no UK law that gives you direct ownership of your own likeness. A consultation on fixing that is due this summer. Until it lands, protecting your face and voice is down to contracts and vigilance, not statute.
What actually changed in February
Section 138 of the Data (Use and Access) Act 2025 amended the Sexual Offences Act 2003 so that "intimate image" now explicitly covers AI-generated and digitally manipulated content, not just photographs and film. It's now an offence to create a non-consensual intimate image of an identifiable adult using AI, and it's an offence to request that someone else create one, even if the image never actually gets made. That second part matters: prompting a tool and asking a person to run it for you are both caught.
This closed a real gap. Before it, someone could feed your public photos into a face-swap tool and the law had almost nothing to say about the output, because no camera had captured anything real. Worth knowing if you deal with this kind of abuse yourself, or if a follower reports it to you - there's now an actual offence to point to, not just a platform report button.
What it doesn't touch
The February change is narrowly about non-consensual sexual content. It says nothing about the much more common commercial problem: someone cloning your voice to read out a script for a product you've never used, or generating a video of "you" recommending a supplement, a crypto scheme, or a competitor's service. That's not an intimate image offence. It might be fraud, it might be trade mark or passing-off territory if your name is used as a brand, it might breach platform terms - but none of those routes give you a clean, direct claim that says "this is my face, my voice, my choice."
The House of Lords Communications and Digital Committee said as much in its March 2026 report on AI and copyright: performers and creators are exposed because the UK has no standalone personality right or voice-likeness right. The Committee's recommendation is a new statutory protection - somewhere between a narrow "digital replica right" for AI-generated fakes and a broader personality right covering your identity more generally - and the government has committed to a consultation this summer to work out which. As of now, that consultation hasn't concluded. Nothing has changed on the ground yet.
Why this matters more once you have an audience
A cloned voice or face is a different order of problem for a working creator than for a private individual. Your face and voice are how people recognise you, and recognition is the actual asset a brand is paying for when it books you. If a convincing fake of you is circulating - even a bad one - it muddies exactly the thing your income depends on: audiences trusting that when they see or hear you, it's actually you.
It's also the kind of harm that's hard to price and slow to notice. A cloned ad might run for weeks in a corner of the internet you don't check before anyone flags it to you. There's no equivalent of a copyright strike system for "someone is using an AI version of my voice."
What to do while the law catches up
None of this is legal advice, and it isn't a reason to panic. It's a reason to close the gaps you can close yourself before a statutory right exists to lean on.
- Put likeness clauses in brand and agency contracts now. Don't rely on generic usage-rights wording written before generative AI was a consideration. Specify whether a brand or platform can use AI to generate content featuring your face, voice, or "style" at all, and for how long any approved usage lasts.
- Search for yourself periodically. A reverse image search on your own thumbnails and a text search for your name plus "AI voice" or "deepfake" takes a few minutes and is currently the closest thing to monitoring that exists.
- Know the reporting routes you actually have. Platform terms generally prohibit impersonation and non-consensual synthetic media even where UK law doesn't yet give you a personal right, so a takedown report is usually faster than any legal route regardless. Under the Online Safety Act, platforms face fines of up to 10% of global turnover for failing to act on this kind of content, which gives your report more teeth than it used to have.
- Document your baseline. Keep dated originals of your key content and any verification marks you use. If a personality right does arrive, having a clear record of what's genuinely yours will matter for enforcing it.
None of this is legal advice - it's general information about a fast-moving area of UK law. If you're dealing with an actual case of cloned or misused likeness, or negotiating a contract clause around AI usage rights, that's a conversation for a media lawyer, not a blog post.
Where this fits
Protecting how you're actually represented online is close to the core of what a management agency is for, even before the law formally recognises it as a right. CFBM Management Services handles the repurposing and channel management work day to day, which includes keeping an eye on how a creator's content and identity show up across platforms - one less thing to monitor alone.
Get these as they go out
Two a week, on what actually changed and what to do about it. No pitch.
We only use it to send the posts. Unsubscribe any time.