Acceptable use policy

Scope and current status

Ocean Studio operates Lobby Duet AI. This policy forms part of the Terms. It covers website material and any authorized portraits, reference video, audio, instructions and results processed through the service. Public accounts, uploads, generation and purchases remain closed. Publishing this policy does not mean a staffed public moderation system is already operating.

Six prohibited content categories

1. Pornography / NSFW: sexually explicit acts, sexualized nudity, intimate imagery or sexual exploitation. Adult consent does not permit these uses.
2. Violence / gore: graphic injuries, torture, cruelty, threats, glorification of violence or instructions encouraging harm.
3. Hate speech: hateful, dehumanizing or discriminatory attacks on protected groups and incitement to hatred or violence.
4. Child-unsafe content / CSAM: child sexual abuse material, sexualization, exploitation or grooming of minors, and content endangering children, including AI-generated or manipulated depictions. The initial generation scope is consenting adults; do not submit minors' portraits.
5. Deepfakes / impersonation: identity replacement without informed permission, deceptive impersonation, fabricated endorsements or representing synthetic footage as an authentic recording of another person. Disclosed fictional adult entertainment does not permit deception, harassment or prohibited intimate content.
6. Copyright / trademark infringement: use of portraits, performances, footage, music, brands or other protected material without required rights; misleading sponsorship or affiliation claims are also prohibited.

Terrorism, extremist recruitment, assistance with mass violence or weapons of mass destruction, incitement to genocide, fraud, doxxing, harassment, self-harm encouragement and attempts to evade safety or access controls are also prohibited. Severe prohibitions apply to synthetic material too. Both adults must consent, and outputs must not be represented as proof of a real event.

Platform moderation — required before launch

Public generation is closed. The experimental model request enables the provider's safety checker, but that alone is not a verified platform-side review system. A responsible reviewer has not yet been assigned. Before generation opens, we must implement and verify this process:

• Before generation: check adult eligibility, portrait consent, inputs, reference footage, audio and instructions against these rules. Prohibited or unresolved requests must not reach the generator.
• Before delivery: screen the complete output, including visuals, identity use, audio and generated text; hold flagged or uncertain results for human review and do not deliver unchecked raw model output.
• Review and records: assign a responsible reviewer and escalation owner, record screening and decisions, and actively maintain blocking rules. A one-time keyword list is insufficient.
• If public sharing opens: review published work and provide removal controls; private uploads must not automatically become examples.

Actual screening tools, reviewer coverage and data handling will be disclosed when implemented. We do not claim that Waffo Prompt Sift, OpenAI Moderation, automated video screening or a 24-hour review team is connected. A text-only filter would not establish that portraits and full videos are checked.

Reporting and review

Email Hello@lobbyduet.com with subject Content report — Lobby Duet AI. Include the URL or task identifier if available, the suspected category, your relationship to the material and a safe description. Do not send copies of CSAM or other illegal imagery, private portraits or identity documents. A URL or identifier and explanation can start a review.

We assess the available evidence against this policy; a complaint alone does not establish a violation. Severe child-safety, terrorism and immediate-harm concerns take priority. Review may include limited follow-up information, restriction of material under our control, a recorded decision and an outcome notification where lawful and practical. Reporter information is disclosed only as necessary for review, response or a lawful obligation. Deliberately abusive reporting is prohibited.

For an immediate threat, contact local emergency authorities. Email is not an emergency service. We do not promise automated acknowledgements, continuous staffing or a fixed response time that has not been established.

Enforcement

For the public introduction website, we can correct or remove material we publish and restrict abusive traffic. Where required by law, we may preserve limited evidence and cooperate with competent authorities.

Before accounts or generation open, enforcement must support rejecting prohibited requests, withholding or removing content under our control, warnings, access restrictions, temporary suspension and permanent termination for serious or repeated misuse. Suspected CSAM, terrorism or serious unlawful activity requires urgent escalation and legally required reporting. These are required launch capabilities; no public account-ban or output-deletion system is represented as already operational.

Action must be proportionate to evidence and severity. We cannot promise deletion of copies held by unrelated third parties. Refunds, credit adjustments and retained evidence follow applicable law and the Refund Policy and Privacy Policy. A report does not automatically forfeit purchased credits.

Appeals and updates

Reply to the same email with subject Appeal — Lobby Duet AI, the decision/content identifier and your reasons. We will consider additional evidence and correct a mistaken decision where supported. We do not promise an independent review team or fixed appeal deadline before that process is established.

Policy updates will be reflected in these shared pages. New features remain subject to these standards and actual release checks.

Terms · Privacy · AI models · Contact