Acceptable Use Policy
What you may not upload or generate. The list is exhaustive and applies to both your uploads and the output.
Version v1.1 · in force from 22 September 2026
People
1. Images or video of minors, in any context and for any purpose.
2. Sexual or sexualised content, including any sexualising manipulation of images of real people.
3. Deepfakes of real people without documented consent: generating, manipulating or animating the face, voice, body or likeness of an existing, identifiable person — including public figures, entertainers, politicians and athletes — without their authorisation.
4. Content depicting real people saying or doing things they did not say or do, where it is capable of misleading.
5. Content that impersonates a person, a brand or a company.
There is no exception to points 1, 2 and 3. There is no “proceed anyway” button, and there will not be one.
Intellectual property
6. Copyright-protected characters, works and settings — including characters from animation, comics, film and video games — where the output is intended for commercial, advertising, promotional or resale use.
Put usefully: generating an image inspired by a style for personal use is a different thing from generating a protected character in order to sell it or advertise with it. The prohibition targets the second case.
7. Third-party trademarks, logos, packaging and trade dress, and content imitating an existing advertising campaign in a confusing way.
8. Uploaded material you do not hold the rights to.
9. Using output to train competing AI models, or to deliberately reproduce the works those models were trained on. This is not our choice: it is a limit the model providers impose on us, and it passes on to you (see Terms, clause 3).
Content
10. Violent, hateful or discriminatory content on grounds of race, ethnicity, religion, nationality, gender, sexual orientation or disability; content inciting crime, self-harm or terrorism.
11. Disinformation about health, elections, public safety and civil protection.
12. Content intended for fraud, phishing or scams.
Consequences
Content removal, account suspension, termination without refund in serious cases, and reporting to the authorities where required.
Repeat infringement: first established breach, formal warning and removal; second, 30-day suspension; third, permanent termination without refund.