AI Content Policy
How generated material is produced, labelled and permitted to be used.
Last updated 23 Aug 2026
Art City AI generates images, video, voice and music. This policy says how that material is produced, how it is labelled, and what you may do with it.
How output is produced
Jobs run on a mix of Art City’s own fine-tuned models and third-party providers. Which provider handled a job is recorded against that job and visible to you.
Placeholders are labelled
Where a provider is unavailable, the platform may return an Art City stand-in rather than failing. That result is marked as a placeholder everywhere it appears, and delivery QC treats an unreplaced placeholder as a failure. A placeholder must never reach a client as finished work.
Synthetic voice and likeness
- Voice cloning requires a documented consent record against the person cloned.
- Generating a recognisable real person's likeness requires their permission, recorded in the Rights module.
- Neither is permitted for political content or for anything designed to be mistaken for a real recording of a real event.
Disclosure
Where a broadcaster, platform or regulator requires AI-generated material to be disclosed, that obligation is yours as the publisher. The delivery manifest records which deliverables contain generated material so that the disclosure can be made accurately.
Training
Your project content is not used to train models served to other customers unless you have signed a separate agreement saying so. Datasets used for Art City’s own fine-tuning carry a rights record naming the copyright owner and the permitted uses.
