Managed AI does not generate three categories of content. All other content is permitted.
This policy sets out the content that Orihai's managed AI service will not generate. It does not restrict genre, tone or subject matter beyond the categories listed.
01Scope
This policy applies only to managed AI.
Use of Orihai with your own provider key or with a model on your own computer is governed by that provider's terms and by applicable law. Orihai does not moderate such use, and no content is sent to Orihai's servers.
02Prohibited content
The following categories are prohibited in every request and in every mode. No setting unlocks them.
- Sexual content involving minors. No exemption applies for fictional content.
- Sexual content depicting identifiable real people.
- Content that is illegal to produce or distribute, whether or not it is fictional.
03Minors
All characters in a sexual context must be adults. Indications of age, appearance or setting that suggest otherwise cause the request to be refused, as do ambiguous cases.
This check applies to imported character cards as well as to typed input. An age stated in a character card is treated in the same way as an age stated in a message.
04Permitted content
Sexual content between adults, violence, dark themes and distressing subject matter are not prohibited. Fictional non-consent between adults is permitted.
Adult content is not the purpose of Orihai, and nothing in the default experience is oriented around it.
05Locked and unlocked scenarios
Scenarios are locked by default. In a locked scenario, managed AI is instructed to keep the writing non-explicit.
To unlock a scenario, you must affirm on your account that you are 18 or older. In an unlocked scenario, only the prohibited categories in clause 02 apply.
The affirmation is held on the account. Unlocking is a per-scenario setting and cannot grant more than the account permits.
06Refused requests
When a request is refused, generation stops and the rule that applied is shown. Orihai records which rule applied, against the generation and against the account. The text of the request is not recorded.
07Provider filtering
The model used by managed AI applies its own content filtering, which is in places stricter than this policy. It may refuse or stop generating content that this policy permits.
Such refusals are failed generations, not applications of this policy. Provider filtering does not apply to use of Orihai with your own provider key or with a local model.
08Changes
This policy took effect on the date shown at the top of this page. It forms part of the terms of service. Changes to the prohibited categories will be notified to the email address on your account before they take effect.