Vibelens

Acceptable Use Policy — AI-Generated Content

Last updated: 2026-10-19

Vibelens, operated by zexuan wang, applies this policy to prompts, permitted scene/style references, and generated images or videos.

1. Scope

All submitted, generated and privately stored content is covered. This service provides fictional scene creation; it has no public gallery, face-swap, nudify, voice-cloning or real-person impersonation feature.

2. Content standards and prohibited categories

The following six categories are prohibited individually:

1. Pornography / NSFW: nudity, explicit sexual material and sexualized depictions.
2. Violence / gore: graphic injury, torture, terrorism, weapons-of-mass-destruction instructions, or promotion of self-harm.
3. Hate speech: dehumanization, incitement to hatred, genocide or violence against protected groups.
4. Child-unsafe content (CSAM): child sexual abuse material or any sexual content involving minors. Zero tolerance.
5. Deepfakes / impersonation of real people: face swaps, nudification, fabricated intimate scenes (including kissing/hugging composites), or fabricated statements/actions, with or without consent.
6. Copyright / trademark infringement: unauthorized use of protected characters, brands, logos or other protected material.

2.1. Absolutely prohibited

CSAM, terrorism/extremism or mass violence, weapons-of-mass-destruction instructions, incitement to genocide/ethnic hatred and real-person deepfakes are blocked. Serious unlawful conduct may be referred to appropriate authorities.

2.2. Restricted and misleading requests

Ambiguous requests are held or refused. This service has no adult-content exception or age-gated NSFW tier. Deceptive political content and misleading medical, legal or financial claims are not permitted.

2.3. Prohibited conduct

Do not spread misinformation, impersonate individuals/brands, violate privacy or intellectual property, pass AI output off as human-made, generate spam, jailbreak moderation, bypass payment or abuse free trials.

2.4. Sources of uploaded material

Only scene/style reference templates accept images. Do not upload people, faces or voice recordings. References must be your own or material you are authorized to use. Uploading another person's photos or voice without explicit permission is prohibited. Permission never makes face swaps, nudification, intimate-scene synthesis or fabricated statements/actions acceptable.

3. Content moderation

3.1. Automated review

Before generation we check a maintained D1 blocklist and, when enabled, Waffo Prompt Sift. A block is refused; a pending review is held without charging or starting generation. After generation and before display or R2 storage, images are assessed with Cloudflare Workers AI Llama 3.2 11B Vision Instruct against the six categories above. Available video covers/first frames receive the same check; a still image cannot certify every video frame. If no cover/first frame is available, we rely on provider moderation only and record that coverage limitation. Unsafe or unreviewable image results are withheld and eligible credits returned. Review bypass is restricted to local development.

3.2. Human review

The operator reviews held requests and substantiated user reports, records decisions and may remove violating stored outputs. Approval of a held request does not charge or generate automatically: resubmit for fresh automatic checks.

3.3. Periodic audits

The operator maintains prohibited terms, reviews moderation logs and provider-only video cases, and adjusts filters based on reports. Safety audits are performed by the operator; this service does not claim a 24/7 review team or an independent audit.

3.4. Severity

L1 critical: child safety, terrorism or mass violence; immediate block and priority review. L2 high: serious prohibited content; withhold/remove and review. L3 medium: other policy violations; refuse/remove with explanation. L4 low: ambiguous or correctable requests; ask for changes.

4. Reporting

Report prohibited content, impersonation, infringement or child-safety concerns with a job reference and explanation. Reports are assessed objectively; we protect reporter identity to the extent permitted by law and review malicious reports as abuse. Do not send unlawful imagery by email.

5. Reporting channels

The operator's support email is also the safety/reporting channel. Email wzexuan707@gmail.com.

6. Response and processing

The operator triages reports by severity, examines logs and available content, records a decision, applies available measures and responds through the support channel. Critical child-safety and terrorism reports receive priority. No automated acknowledgement or guaranteed response time is claimed; the operator must publish realistic staffed timelines before offering a guaranteed SLA.

7. Enforcement

Available measures are request refusal, temporary review holds, warnings, suspension of generation access for a violating browser session, withholding unsafe outputs, removal of stored violating content and revocation of unused credits on an approved payment refund. Severe suspected unlawful conduct may be reported to authorities. This no-login service does not claim permanent identity bans or automatic legal enforcement.

8. Appeals

Contact support with the job reference and reasons for review. The operator records the outcome and may request a revised prompt. Escalation to a separate reviewer is arranged where available; no automatic reversal bypasses safety checks.

9. Transparency and updates

Model and moderation coverage are disclosed on this site. Material policy changes are published here. There is no public feed; any future gallery requires additional publication-stage scanning before launch.

10. Contact

Operator: zexuan wang. Safety, reports and appeals: Email wzexuan707@gmail.com. See also our terms, privacy policy and pricing.