Synthetic,
and says so.

AI-generated video is now convincing enough to be useful and convincing enough to be misused. The useful part is speed and volume: thirty opening hooks tested before a single creator is hired, one script localised into five languages in an afternoon, an explainer that can be updated when the product changes without reshooting. Those are real savings and we use them.

The misuse is presenting a synthetic person as a real customer. An AI presenter saying "I have used this every day for a month" is a testimonial from someone who does not exist, and consumer protection law in the EU and the US treats it accordingly. The format can be the same; what changes is the script, which speaks about the product rather than inventing an experience of it.

Disclosure is part of the build. The EU AI Act requires synthetic video and audio of realistic people to be identified as AI-generated, with those obligations applying from August 2026, and the major platforms require or apply their own labels. We label the content and set the platform disclosure at upload, rather than gambling that nobody notices.

Likeness is the other line. Stock AI presenters are licensed from the provider for commercial use, and we check those terms. A custom avatar or cloned voice of a real person — your founder, a spokesperson, a creator — is only made with that person's written consent and a contract stating where it may be used.

What you get

• Hook and angle variants generated at volume for testing
• Licensed AI presenters and voices, with commercial terms checked
• Localised versions of a script in other languages
• AI disclosure in the content and in the platform settings
• Custom avatars or voices only with written consent

Not included

• Media buying and ad account management
• Creator fees and product samples, quoted per brief
• Fake reviews or synthetic “customer” testimonials