Policy

Trust & Safety

Effective date: July 20, 2026

Our safeguards focus on informed consent, accountable generation, private-by-default assets, explicit publishing, and decisive action against deceptive or harmful synthetic media.

1. Our commitment

Skyrasia is an AI creation platform. Synthetic media should expand what people can make, not who they can deceive. This policy explains our consent requirements, publication safeguards, reporting routes, and enforcement practices. It operates alongside the Terms of Service and Privacy Policy; the Terms govern where documents differ.

2. Consent-first voice and likeness

  • Only clone or reproduce a voice or likeness you own or have informed, written permission to use.
  • The same rule applies to reference images, avatar training, lip synchronisation, and synthetic depictions of real people.
  • We may request evidence of permission and remove models or content when evidence is missing or inadequate.
  • Fraud, political manipulation, non-consensual sexual content, and harmful deception remain prohibited regardless of a claimed consent.

3. Provenance and accountability

Skyrasia records generation, project, model, connected-application, credit, and asset events needed for delivery, security, billing, and abuse investigation. Provider-level provenance metadata or watermarks may be present where supported. We do not currently represent that every output carries a proprietary imperceptible watermark. Creators must follow applicable synthetic-media disclosure and labelling rules and must not remove required provenance signals.

4. Rights attestation for Discover

Discover is opt-in. Before publishing, a creator must affirm that they control the rights needed to distribute the selected video and everything in it, and that it follows this policy. We associate publication with the creator, project, selected asset, time, and policy version. We may reject, unlist, age-restrict, or remove a video in response to policy, safety, or rights concerns.

5. Prohibited uses

  • Impersonation of a real person without express authorisation, including public figures.
  • Fraud, scams, identity theft, blackmail, or social engineering using synthetic media.
  • Political manipulation or fabricated candidate, official, election, or public-policy media presented as authentic.
  • Misinformation deliberately designed to deceive audiences.
  • Harassment, threats, stalking, bullying, or defamation.
  • Sexual content involving minors, or non-consensual sexual content involving any person.
  • Uploading a voice or likeness without the owner’s knowledge and informed permission.
  • Publishing material without distribution rights.

6. Reporting

Anyone can report abuse, impersonation, non-consensual media, or a rights concern to hello@skyrasia.com. Include a link or copy of the content, a clear description, and, where relevant, your identity or authority to act for the affected person. Copyright notices should identify the protected work, the allegedly infringing video, contact information, and a good-faith statement that the use is unauthorised.

7. Enforcement

  • Rapid disabling or removal of content where a violation is verified or a credible complaint requires investigation.
  • Removal of unauthorised cloned voices, replicas, and related models.
  • Warnings, tool restrictions, suspension, or permanent termination based on severity and history.
  • Correction or withholding of credits or other benefits tied to fraud or violating activity.
  • Restriction or termination of repeat infringers.
  • Escalation of suspected illegal activity and cooperation with lawful requests from authorities.

8. Platform safeguards

  • The Platform is limited to people aged 18 or older.
  • Private assets remain workspace-scoped; Discover requires explicit publication.
  • Generation and connected-application activity is logged for security, billing, and investigation.
  • Paid provider calls use an auditable credit ledger and retry-safe job identities.
  • Payments are handled by established processors; Skyrasia does not store full card details.
  • Private identifiable content is not used to train general-purpose AI models without explicit consent.

9. Partners and regulators

Our operating framework is consent-first: users are responsible for permission to use voices and likenesses; publication is explicit and attributable; prohibited-use rules ban impersonation, fraud, and harmful deception; private and published content are separated; and reports feed a documented review and enforcement process. Partners, processors, and regulators may request additional control information at hello@skyrasia.com.

10. Contact

Company: BNX Creative Labs Ltd (RC 9601091), operating Skyrasia. Abuse reports, takedown requests, compliance inquiries, and safety questions: hello@skyrasia.com. Website: https://skyrasia.com.

← Back to home