Safety, Likeness & Consent
Effective November 2026
AI Acceptable Use Policy
Mandatory consent rules for voice cloning and human likenesses, prohibited content categories, and upstream model safety compliance.
1. Mandatory Consent for Voice Cloning & Human Likenesses
- Voice Cloning & Speech Synthesis (ElevenLabs & Audio Engines): You represent and warrant that you hold explicit, verifiable permission from the speaker before uploading voice samples, cloning a voice, or generating synthetic speech in a real person's voice.
- Human Likeness, Motion & Lip-Sync (Runway Act-Two, Kling, Veo & Reference Keyframes): You may only upload photographs, video performances, or facial reference sheets of real individuals when you have their documented consent (or valid talent release) for AI synthesis.
2. Strictly Prohibited Content
You may not use RenderHaus Cloud, presigned S3 storage, or scoped MCP tokens to upload, prompt, generate, or distribute:
- CSAM / CSBD: Any child sexual abuse material or sexualized depiction of minors (zero tolerance; results in immediate permanent termination and reporting where required by law).
- Non-Consensual Intimate Imagery (NCII): Sexually explicit or intimate depictions of any real person without their consent.
- Deceptive Impersonation, Fraud & Disinformation: Unlawful impersonation of individuals, executives, public officials, or biometric liveness bypasses intended to defraud, extort, or deceive.
- Self-Harm, Cyberattacks & Unlawful Violence: Content promoting suicide/self-harm, CBRN weapons, malware, or real-world violent extremism.
3. Upstream Provider Safety Compliance & Enforcement
All jobs routed through RenderHaus Cloud remain subject to the safety policies and automated content filters of MiniMax, Runway, Kling, Google, xAI, Luma, OpenAI, Black Forest Labs, ElevenLabs, and fal.ai. Repeated or egregious violations of this policy may result in immediate workspace suspension. Report abuse to owner@vehicledesk.io.