Acceptable Use Policy & Content Moderation Guidelines
Effective Date: October 2026 • Legally binding policy governing all voice synthesis, cloning models, and API endpoints
SynVoice (accessible at https://synvoice.io) operates a zero-tolerance policy against unlawful, deceptive, or abusive synthetic audio generation. This Acceptable Use Policy ("AUP") defines mandatory content standards, explicit censorship boundaries, automated screening mechanisms, and multi-tier enforcement protocols applicable to all registered users, enterprise accounts, and API integrations.
Compliance with this policy is a strict prerequisite for maintaining access to SynVoice services and credit balances. Violations result in instantaneous account termination and forfeiture of purchased credits.
1. Scope & Applicability
This Policy forms an integral, legally binding part of the Terms of Service. It applies to every person or entity who accesses the SynVoice platform, whether through our graphical web interface, browser extensions, mobile views, or developer REST APIs (e.g., /v1/text-to-speech, /v1/text-to-dialogue).
By submitting text prompts, uploading audio reference samples, or calling any SynVoice endpoint, you explicitly represent and warrant that your generated material complies fully with all rules outlined herein.
2. Zero-Tolerance Prohibited Content & Abuse
You are strictly forbidden from submitting prompts, scripts, or audio samples designed to synthesize, distribute, or facilitate any of the following categories of content:
Defamation, Hate Speech & Violence
Any content that incites violence, promotes terrorist organizations, encourages self-harm, harasses individuals, or targets protected classes based on race, ethnicity, religion, sexual orientation, disability, or national origin.
Non-Consensual Voice Cloning & Deepfakes
Cloning or synthesizing the voice of any individual without verifiable affirmative written consent. Unauthorized impersonation of celebrities, public officials, corporate executives, law enforcement officers, or political figures.
Deceptive Media & Election Interference
Fabricated recordings designed to deceive voters, misrepresent polling information, impersonate candidates, sway democratic elections, or distribute disinformation intended to manipulate civil society.
Fraud, Robocall Spam & Biometric Spoofing
Voice phishing ("vishing"), automated robocall spam, banking fraud, caller-ID spoofing, biometric authentication bypass attempts, extortion, or non-consensual financial collection scams.
Sexually Explicit Material & Exploitation
Child Sexual Abuse Material (CSAM) or Child Sexual Exploitation and Abuse (CSAE) in any form. Non-consensual sexually explicit deepfakes, intimate audio, or predatory grooming material.
Intellectual Property & Piracy
Unauthorized commercial reproduction of proprietary audiobook recordings, copyrighted voice performances, trademarked catchphrases, or audio assets where you lack required commercial exploitation licenses.
3. Content Moderation & Active Censorship Architecture
To ensure public safety, legal compliance, and fintech payment partner integrity, SynVoice maintains an active, multi-layer moderation system integrated directly into our generation pipeline:
Layer 1Automated Pre-Generation Text Filtering
Every script submitted via web interface or API undergoes real-time lexical inspection before synthesis models execute. Prohibited strings, terroristic incitement patterns, abusive keywords, and fraudulent scam phrases are automatically rejected with an HTTP 400 CONTENT_POLICY_VIOLATION error code. No inference tokens or audio bytes are generated for flagged prompts.
Layer 2Real-Time Heuristic & Anomaly Detection
Our ingestion infrastructure continuously monitors for burst generation anomalies, repetitive political speech scripts, unauthorized automated scrapers, and malicious prompt injections intended to bypass censorship rules.
Layer 3Cryptographic Audio Fingerprinting & Watermarking
Generated audio outputs include verifiable cryptographic audio metadata markers and file signatures confirming synthesis through SynVoice. This facilitates trace identification and takedown should synthetic media be illegally distributed across public platforms.
Layer 4Human-in-the-Loop Triage & Abuse Auditing
Suspicious generation attempts, user abuse reports, and automated security escalations are reviewed by our dedicated Trust & Safety compliance team for definitive remediation.
4. Voice Cloning Ethics & Mandatory Biometric Consent
Instant Neural Voice Cloning is one of the most powerful capabilities provided by SynVoice. To prevent identity theft and biometric misappropriation, users who utilize voice cloning must satisfy the following strict requirements:
🧬Affirmative Representation of Legal Authority & Biometric Consent
By uploading any voice audio sample (e.g. MP3, WAV, M4A) to SynVoice for cloning, you expressly, affirmatively, and legally represent that:
- You are either the speaker whose voice is being cloned, OR you have received explicit, written, verifiable consent from the voice subject authorizing you to clone and synthesize their voice commercially.
- You possess all necessary copyright licenses for the uploaded audio recording and will not infringe on any third party's right of publicity, trademark, or biometric privacy.
- You will never use the cloned voice profile for deceptive impersonation, automated phone scams, fraud, or defamatory media.
- You agree to indemnify, defend, and hold harmless SynVoice.io and its affiliates from any liability, legal actions, or damages arising out of unauthorized voice sample uploads.
5. Multi-Tier Violation Enforcement Matrix
SynVoice applies a clear, graduated enforcement policy proportional to the severity and frequency of policy breaches:
| Enforcement Level | Trigger Condition | Platform Action Taken | Credit Balance Status |
|---|---|---|---|
| Level 1: Warning & Block | Single automated pre-filter prompt match | Immediate request rejection (HTTP 400). System warning displayed. | Credits unaffected. |
| Level 2: Temporary Freeze | Repeated attempts to bypass filters, burst prompt anomalies | API keys suspended for 24–72 hours. Mandatory compliance review. | Frozen pending investigation. |
| Level 3: Permanent Ban | Non-consensual voice cloning, hate speech, election fraud, scam vishing | Immediate permanent account termination. IP & identity blacklisted. | Forfeited permanently. No refund. |
| Level 4: Legal Escalation | CSAM, extortion, terror threats, criminal fraud | Complete ban, incident dossier compiled, and referral to law enforcement agencies. | Forfeited permanently. |
6. Merchant Processor & Crypto Gateway Compliance
SynVoice integrates with reputable payment gateways (including credit card networks and regulated cryptocurrency payment processors such as Heleket). In alignment with payment processor underwriting guidelines and anti-money laundering (AML) / counter-terrorist financing standards:
- No High-Risk / Prohibited Commerce: SynVoice accounts may not be purchased using stolen payment credentials, mixing services, or illicit funds.
- Non-Refundable Computational Allocation: Because synthetic GPU inference incurs immediate irreversible computational overhead, credits purchased and consumed are non-refundable.
- Zero Chargeback Defense for Banned Violators: Any user whose account is terminated for violating this Acceptable Use Policy explicitly forfeits all claims to credit refunds, reversals, or payment chargebacks.
7. Abuse Reporting, DMCA Takedown & Contact
We encourage the community, voice actors, copyright holders, and regulatory authorities to report any suspected abuse, non-consensual voice cloning, or policy violations. Our Trust & Safety team investigates all reports within 24 hours.
Platform: SynVoice (https://synvoice.io)
Trust & Safety Hotline: support@synvoice.io
General Support Email: cs@synvoice.io
DMCA Copyright Inquiries: support@synvoice.io (Subject: DMCA Takedown Notice)
Need to report an urgent unauthorized voice clone or abusive prompt?
Report Abuse to Trust & Safety→