AI Disclosure & Responsible Use
Effective 20 November 2026
ELARVUM is an AI-native film studio. We believe audiences, clients and collaborators deserve to know how our work is made. This page explains how we use generative AI and the rules we hold ourselves to.
1. How we use AI
We use generative AI tools to create images, video, voices, music and sound. People make the creative decisions: the story and script, visual design, casting of characters, selection of takes, editing, pacing, sound design, mixing and colour grading. Every ELARVUM project is directed by a filmmaker.
2. What on this site is AI-generated
Unless a credit says otherwise, the films, trailers, posters, characters and visual effects shown on this site have been created with generative AI. We say so on every film page, in this disclosure, and in credits or descriptions where we publish our work. The founder's portrait and written text on this site are produced and approved by ELARVUM.
3. People, likeness and consent
- We do not depict a real, identifiable person — their face, body or voice — without their written consent.
- We do not create deceptive deepfakes, political or election-related impersonation, non-consensual intimate imagery, or any sexualised content involving minors.
- Voice cloning is used only with the documented consent of the speaker.
- Fictional characters are original creations. Any resemblance to real persons is coincidental, and we act quickly if someone believes a character resembles them.
4. Music and voices
We create original AI-assisted music, songs and voice work. We do not imitate the voice or distinctive style of a named living artist without permission, and we do not use copyrighted recordings or lyrics we have no licence for.
5. Tools and licences
We work on commercial plans of established generative AI tools and follow their usage policies. Our toolkit changes as technology improves; current tools are listed on our About page. Where a tool embeds provenance metadata or watermarks in its output, we do not remove them.
6. Labelling and the law
Since 2 August 2026, the EU AI Act (Article 50) requires deployers of AI systems to disclose deepfakes — AI-generated or manipulated content that resembles real people, places or events. For evidently artistic, creative or fictional works, this disclosure may be made in a way that does not interrupt the work itself. ELARVUM applies this approach to all audiences, not only in the EU: we disclose AI use clearly in credits, descriptions and on this site, and we follow the AI-labelling rules of the platforms where we publish.
For client work, we agree the labelling approach with each client, in line with the laws of the countries where the content will run.
7. Copyright and ownership
Copyright law for AI-assisted works is evolving and varies between countries. In some jurisdictions, material generated entirely by a machine may not be protected by copyright on its own, while human creative contributions — writing, selection, arrangement, editing and sound — can be. We document the human authorship in our productions. Clients should take their own legal advice where rights are critical.
8. Your data and our models
We never use client materials, inquiries or personal data to train AI models. Client assets are uploaded to tools only as needed for the project and only with tools whose terms we have reviewed.
9. Concerns and reports
If you believe our content uses your likeness, voice or work without permission, or is misleading, email hello@elarvum.com with the subject "AI content concern". We review every report promptly and remove content where appropriate.