A campaign should strengthen a brand without turning a model identity into a vehicle for abuse. Shoodio checks submitted assets and the intended shoot context before generation. The current public catalogue contains adult model identities only.
Protection is part of the production workflow—not an afterthought.Clear boundaries help serious fashion teams work faster. They protect reusable talent, customer brands and the channels where finished campaigns appear.
Every model in the Shoodio catalogue is synthetic and AI-generated. No real person is depicted, and no catalogue identity is derived from a photograph of someone. These persistent identities may not be sexualised, defamed, placed in propaganda or used outside their licensed production context.
The public catalogue and customer-facing model selection contain only verified adult model identities.
Custom locations must be credible fashion settings with usable perspective and space. Hate material, propaganda and recognisable bystanders are not accepted.
When something is restricted, the user receives a practical reason, the affected component and the available next step—without exposing exploitable detection details.
Checks are performed where they are useful: once when a custom asset enters Shoodio and again when the exact shoot is submitted.
Uploaded fashion products, accessories, print-on-demand items and locations are classified against platform, adult-only and product eligibility rules.
The decision is bound to the exact file. Replacing or materially changing the file requires a new assessment; users cannot edit the safety result.
At Shoot, Shoodio evaluates the selected model, products, location, pose, requested action and intended prompt as one production context.
Approved shoots proceed. Restricted shoots identify what must change. The generation provider may still apply its own non-overridable safeguards.
Context matters, but some categories are incompatible with a platform built around trusted model identities.
An unsuitable location is not treated like deliberate extremist abuse. Enforcement follows the severity and confidence of the finding.
| Level | What it means | What happens |
|---|---|---|
| T1 · Restriction | Unclear, unsuitable or context-limited content without a confirmed abuse finding. | The asset or shoot is declined or limited. No strike is added. Enhanced Review may be available. |
| T2 · Violation | A clear policy violation that is serious but not in a zero-tolerance category. | The request is blocked and one active strike is recorded. Three active strikes place the account on Security Hold. |
| T3 · Zero tolerance | Severe abuse such as Nazi propaganda, child sexual exploitation, terrorist recruitment or targeted violent threats. | Immediate Security Hold and specialist review for permanent exclusion. This route has no automated Enhanced Review. |
T2 strikes expire after 12 months. Duplicate or lightly modified copies of the same asset do not create extra strikes, reset limits or bypass a decision.
For eligible T1 and T2 asset decisions, the account owner can request up to three independent enhanced reviews within seven days.
Explain the intended fashion use and confirm that you hold the necessary rights.
After eligibility is confirmed, one credit is reserved and deducted at submission. It is kept only after a substantive review; a terminal technical failure or service refusal is refunded automatically and consumes no completed attempt.
Each review evaluates the unchanged asset, the original decision, the current policy and the submitted statement.
Repeated identical statements or a modified file do not count as a meaningful new review.
Enhanced Review cannot override zero-tolerance rules, missing rights or identity restrictions. Account holds and bans also have a free human correction channel.
Shoodio model identities are production partners, not generic raw material. Our controls are designed to preserve their long-term commercial and personal integrity.
A persistent model identity cannot be used to imply political endorsement, hate, harassment or a statement the represented person never made.
Cloning, training on, extracting or redistributing a Shoodio model identity outside the licensed service is prohibited.
Customers must own or license their uploaded products, trademarks, locations and other material. A successful technical check does not create those rights.
Policy decisions are versioned and auditable. Restricted admin access and recorded revisions prevent silent changes to enforcement history.
Shoodio assesses submitted assets and the intended shoot context before generation. No automated system is infallible, and approval is not a legal opinion or a guarantee that an output is suitable for every market, audience or advertising channel.
After an approved generation completes, Shoodio does not automatically inspect every resulting image or video again. Customers must review the final asset before publishing it and remain responsible for claims, clearances and local law.
We do not infer a user’s political or religious beliefs from an uploaded motif. We classify the content and its requested use solely to operate the service safely.
Initial public release: asset checks, whole-shoot review, an adult-only public catalogue boundary, enforcement and Enhanced Review.
