We filter harmful content at the generation stage, not just after. If content gets flagged, we block it and log it. We don't generate content for minors that we wouldn't generate for adults. We test our safety systems regularly and publish the results.
Skylina is committed to safe, responsible AI use. We implement: (a) pre-generation content filtering on prompts; (b) post-generation quality checks; (c) human review of flagged content; (d) continuous safety system improvement.
Safety filtering operates at multiple stages: (1) Prompt analysis: User input is checked for prohibited content patterns before generation; (2) Output validation: Generated content is scanned for safety violations; (3) Human escalation: Ambiguous cases are reviewed by human moderators; (4) Account monitoring: Users with pattern of safety violations are flagged for review.
Accounts verified as ages 13-17 have additional restrictions: (a) no public sharing of generated content; (b) no access to community or collaboration features visible to non-users; (c) no AI chat integration for adult topics. Parental consent verification is required for accounts under 16 in applicable jurisdictions.
If you encounter unsafe content generated by or hosted on Skylina, report it immediately at safety@skylina.polsia.app. We respond within 24 hours for critical safety issues.
We are committed to monitoring AI outputs for bias. If you believe an output was discriminatory or unfair, report it via our safety team. We document and address bias reports in our quarterly bias testing reports.
AI-generated images and video include invisible C2PA-compliant metadata marking them as AI-produced. This helps distinguish authentic human-created content from AI-generated material.
We maintain a register of serious incidents and safety system failures as required by the EU AI Act Article 11. Incidents are reviewed within 72 hours and remediation steps are documented.