// AI NEWS · AI ETHICS

British Singer IAMDDB Says an AI-Generated Propaganda Video Used Her Likeness Without Consent

The Manchester artist posted that a politically charged AI video copied her identity for "something I 100% did not consent to" — the latest case of a real person's face turning up in synthetic content they never agreed to appear in.

2026-10-17 · by Moe Ameen

What happened

In mid-October 2026, British singer and rapper IAMDDB — the stage name of Manchester artist Diana De Brito, who is of Angolan and Portuguese descent — posted on X that an AI-generated video had used her likeness without her permission. "Hiya yeah so — why did you steal my identity for something I 100% did not consent to?" she wrote, following up that she was "so baffled right now." The clip was politically charged: it circulated as pro-Israel messaging tied to an AI-generated song, and IAMDDB has no connection to the campaign.

According to a community note and posts attached to the thread, the video was built around an AI-generated track credited to a creator posting as Eyal Be, with "GenerativeAI" listed as composer, and the footage appears to draw IAMDDB's likeness from a fashion photoshoot she had done for the brand BALR. In other words, a real commercial image of a real artist was apparently used as source material to animate a synthetic figure delivering a message she never recorded or endorsed. As of writing, the case rests on IAMDDB's own posts and the community note rather than any statement from the video's makers or reporting from a major outlet, so specifics beyond her objection should be treated as developing.

The episode is a clean example of a pattern that keeps recurring in 2026: AI video tools make it trivial to lift a recognizable face from existing photos and drop it into content the person never agreed to — advertising, political messaging, or outright propaganda. It is the same consent-and-likeness problem now driving platform responses such as YouTube's and TikTok's likeness-detection tools, only here the harm is reputational and political rather than commercial.

Why it matters for creators

  • Any public face is now raw material. A single licensed photoshoot image was reportedly enough to generate a video putting words and a political stance into an artist's mouth — the barrier to misusing a likeness has collapsed.
  • Disclosure is not consent. Even a video that is labeled or community-noted as AI-generated does nothing to fix the underlying problem that the person depicted never agreed to appear in it.
  • Detection is reactive, not preventive. Likeness-detection tools let you find and report a deepfake after it spreads; they do not stop one from being made, which is why creators are moving to assets built on faces they actually control.
  • The reputational stakes are real for creators. Being inserted into partisan propaganda can alienate an audience and damage brand deals in ways a takedown weeks later cannot undo.
  • It sharpens the honest line for AI video: the only likeness you should ever build synthetic content around is one you own or are clearly licensed to use — your own face, a team member who consented, or a properly licensed persona.

How to act on this with Kompozy

The lesson creators are taking from cases like this is not "avoid AI video" — it is "only generate video around a likeness you actually control." That is exactly the model [Kompozy](/) is built on. Its avatar formats — [Persona Shorts](/glossary/persona-shorts) and Persona HeyGen — render from an AI Influencer persona you create from your own face, with Gemini face-lock keeping that persona consistent across every video, and a [Persona Brief](/glossary/persona-brief) governing the voice and the words. Nothing is scraped from a stranger's photoshoot; the face in the output is one you enrolled and own, so the consent question is answered by construction rather than after a takedown request. For the full framework on whose likeness you can legally build on, see our guides on [AI avatar likeness consent](/guides/ai-avatar-likeness-consent) and [unauthorized AI likeness ads](/guides/unauthorized-ai-likeness-ads).

Kompozy is not a deepfake detector — finding and reporting a synthetic video of your face is what platform tools like YouTube's and TikTok's likeness detection are for. What Kompozy does is let you own the production side cleanly: build a persona from a face you consent to, generate a captioned avatar video, a brand-exact [Carousel](/glossary/hyperframes), Quote Graphics, a Blog Article, or an Email Newsletter from one source, and let [Autopilot](/glossary/autopilot) schedule and publish the set across the eight social platforms plus blog and email behind a per-post review. If you are also worried about your likeness being misused elsewhere, our guide on [protecting your likeness from AI deepfakes](/guides/protect-your-likeness-from-ai-deepfakes) covers the detection-and-response side the platforms now provide.

Quick takeaways

  • In mid-October 2026, British singer IAMDDB (Diana De Brito) said on X that an AI-generated, politically charged video used her likeness without consent.
  • Per an attached community note, the video was built around an AI-generated pro-Israel track and appears to have drawn her likeness from a BALR. fashion photoshoot.
  • The case rests on her own posts and the community note — no statement from the makers or major-outlet reporting had landed at the time of writing.
  • It is the same consent-and-likeness problem driving YouTube's and TikTok's likeness-detection rollouts: AI makes lifting a real face into synthetic content trivial.
  • The defensible path for creators is to generate video only around a likeness they own or license — which is how Kompozy's persona/avatar formats are designed.

Frequently asked questions

What did IAMDDB say happened?

In mid-October 2026, IAMDDB — British singer Diana De Brito — posted on X that an AI-generated video had used her identity without her consent, asking why her identity was stolen "for something I 100% did not consent to." The clip circulated as pro-Israel messaging she has no connection to. The account rests on her posts and a community note rather than a statement from the video's makers.

Where did the AI video reportedly get her likeness?

According to a community note attached to the thread, the footage appears to draw IAMDDB's likeness from a fashion photoshoot she did for the brand BALR., repurposed into an AI-generated video built around a synthetic track credited to a creator posting as Eyal Be, with "GenerativeAI" listed as composer. Those specifics are unverified by major outlets and should be treated as developing.

Is making an AI video of someone without consent allowed?

It depends on jurisdiction and platform, but using a real person's likeness in synthetic content without permission raises serious consent, publicity-rights, and defamation concerns — and platforms including YouTube and TikTok now run likeness-detection tools to find and remove unauthorized AI depictions. Disclosing that a video is AI-generated does not substitute for the depicted person's consent.

How can a creator make AI avatar videos without this risk?

Only build synthetic video around a likeness you own or are licensed to use. A tool like Kompozy generates its avatar videos from an AI Influencer persona you create from your own face — Gemini face-lock keeps it consistent and a Persona Brief governs the voice — so the face in the output is one you enrolled and control, not a scraped image of someone else.

Related news

← All AI news · Get started →