Acceptable Use Policy

Velluto Last updated: 4 июня 2026 г.

This Acceptable Use Policy (the "AUP") sets out the rules that govern your use of Velluto (the "Services"). It is part of the Terms of Service. By using the Services you agree to comply with the AUP.

Violations may result in immediate suspension or termination of your Account, deletion of your Content and Voice Models, forfeiture of fees, and where appropriate, reporting to law-enforcement authorities.

1. General prohibitions

You may not use the Services to:

1.1. Violate the law. Engage in any activity that violates any applicable national, state, or local law, regulation, or sanctions regime.

1.2. Infringe rights. Infringe, misappropriate, or violate any intellectual-property right, right of publicity, right of privacy, or other right of any third party.

1.3. Defraud or deceive. Engage in fraud, phishing, scams, identity theft, social engineering, or any activity that deceives another person to obtain money, credentials, personal data, or anything of value.

1.4. Harass or harm. Threaten, harass, stalk, defame, or incite violence against any person or group, including on the basis of race, ethnicity, national origin, religion, sexual orientation, gender identity, disability, or age.

1.5. Distribute malware. Use the Services to create or distribute malware, viruses, worms, spyware, ransomware, or any other malicious code.

1.6. Compromise security. Attempt to gain unauthorised access to the Services, other accounts, or any computer system; bypass authentication, rate limits, or technical access controls; probe, scan, or test the vulnerability of the Services without our prior written consent.

1.7. Scrape or automate. Use any robot, scraper, crawler, or other automated means to access the Services, other than the documented API and within published rate limits.

1.8. Interfere with the Services. Disrupt, degrade, or place an unreasonable load on the Services or the networks supporting them, including by denial-of-service attacks or distributed denial-of-service attacks.

1.9. Resell or repackage. Resell, rent, lease, or sublicense the Services or generated Output as a stand-alone product that replicates the core functionality of the Services.

1.10. Train competing models. Use the Services or Output to train, develop, or improve any artificial-intelligence model or product that competes with Velluto.

2. Voice cloning — prohibited voices and uses

The following are absolutely prohibited and are grounds for immediate, irreversible termination of your Account.

2.1. Prohibited voices

You may not create or attempt to create a Voice Model of:

(a) Any individual without their explicit, prior, written consent — including friends, family members, ex-partners, colleagues, or persons you know personally;

(b) Any minor (any person under 18 years of age);

(c) Any deceased person without verified authorisation from the rights holder of their estate;

(d) Any current or recent political candidate, elected official, judicial officer, or head of state — regardless of whether you claim to have obtained that person's authorisation. "Recent" means within five (5) years of leaving office or losing election;

(e) Any voice whose owner has revoked consent or whose owner has been notified to us as a "No-Go Voice";

(f) Any voice of a celebrity, artist, athlete, musician, public figure, or any person identifiable by their voice without verified written authorisation;

(g) Any voice for which you cannot, on request, produce written evidence of consent within fourteen (14) days.

2.2. Prohibited purposes

You may not use the Services (whether with a cloned voice, a curated voice, or any other voice) to:

(a) Create non-consensual intimate imagery (NCII) or any sexual, pornographic, or sexually suggestive content depicting a real person without their explicit consent; this prohibition is absolute and has no defence based on artistic merit, satire, or commentary;

(b) Create any sexual content involving a minor, including computer-generated, animated, or simulated content; we report all such content to the National Center for Missing & Exploited Children (NCMEC) and equivalent authorities in your jurisdiction;

(c) Impersonate any real person in a manner intended to deceive others about the identity of the speaker, including for fraud, "vishing" (voice phishing), business-email compromise, romance scams, or "grandparent scams";

(d) Manipulate elections — create, distribute, or facilitate the spread of misleading audio about elections, electoral candidates, election processes, polling, or election outcomes; this prohibition applies regardless of jurisdiction and regardless of whether the manipulated person has consented;

(e) Defame any person by attributing statements they did not make;

(f) Harass, intimidate, or threaten any person;

(g) Generate audio of a real person committing a crime, confessing to a crime, or admitting wrongdoing that they did not in fact commit;

(h) Bypass voice-based authentication systems (banks, telephony, IVR, voice biometrics);

(i) Spread disinformation intended to deceive the public about matters of public concern, including public health, science, or current events;

(j) Promote terrorism, violent extremism, or organised criminal activity.

2.3. Synthetic-media transparency

When the Services are used to produce audio that could reasonably be mistaken for an authentic recording of a real human being, you must clearly disclose to all listeners that the audio is AI-generated. This requirement is consistent with the EU AI Act (Article 50, applicable from 2 August 2026) and is best practice everywhere.

3. Content and topic restrictions

You may not use the Services to generate, host, or distribute:

3.1. Child sexual abuse material (CSAM) or any content that sexualises minors. Absolute prohibition; reported to authorities.

3.2. Non-consensual sexual content depicting any real person.

3.3. Doxxing content — content that publishes personal information of any person with intent to harass.

3.4. Hate speech that incites violence or discrimination against persons based on protected characteristics.

3.5. Detailed instructions for the creation of weapons (firearms, explosives, biological, chemical, nuclear, radiological), the synthesis of controlled substances, or other content that materially assists serious harm.

3.6. Content that promotes self-harm or suicide without harm-reduction framing.

3.7. Spam, unsolicited bulk communications, or content designed for SEO manipulation at scale.

4. High-risk applications

The Services are not designed, certified, or warranted for use in:

4.1. Safety-critical systems (life support, medical devices, aviation, nuclear, weapons control, transportation safety).

4.2. Critical infrastructure operation (power grid, water supply, emergency services).

4.3. Provision of regulated professional advice (medical, legal, financial, psychological) to consumers without a qualified human professional in the loop and without prominent disclaimers that the output is AI-generated and not professional advice.

4.4. Decisions producing legal or similarly significant effects on individuals (credit, employment, education, housing, insurance, immigration), unless a qualified human reviews each decision before it is communicated to the individual.

If you use the Services in any of these contexts you do so at your own risk, you accept full responsibility for the outcomes, and you indemnify us under the Terms.

5. API and rate limits

5.1. Use only the documented API endpoints. Reverse-engineering internal endpoints is prohibited.

5.2. Respect published rate limits. Burst protection is in place; sustained over-limit usage will result in throttling, then suspension.

5.3. Cache responses where appropriate to reduce load.

5.4. Do not share API keys. Each API key is bound to your Account.

5A. Plan limit structure and unlimited plans

5A.1. Plans with an explicit cap

Plan Limit What happens at the cap
Free 10,000 characters / month generation pauses until the next billing period
Personal 50,000 characters / day (~1.5M / month) generation pauses until the start of the next calendar day (00:00 UTC)

Counter resets are automatic. No "plan upgrade" prompts are forced; the live counter is shown in the product UI.

5A.2. Unlimited plans (Pro, Custom Voice)

Pro and Custom Voice carry no hidden volume thresholds — no daily or monthly caps, no throttling based on cumulative consumption. The only limits that apply are:

  • Technical concurrency limits (rate limit on parallel requests and QPS) published in the API documentation. These protect infrastructure from DoS, not against volume.
  • Daily API call limits stated in the plan description (Pro — 1,000 requests / day).

One user, one Account, normal content-production workflows (voiceover, podcast, video) — use as much as you need, no volume counters to watch.

5A.3. What unlimited does NOT permit

Unlimited does not override the following absolute prohibitions (see also Section 8 of the Terms):

  • One Account = one end-user. Sharing access, multi-accounting, and credential pooling are prohibited.
  • No commercial resale of access to the Services, API keys, or Output as part of a stand-alone product that replicates the core functionality of Velluto.
  • No proxy use of the Account — processing of requests initiated by third parties on their own behalf (multi-tenant resale, white-label SaaS wrappers, API proxies with your own pricing layer).
  • No industrial-scale automated extraction — building a parallel corpus of Output or Voice Models in volume disproportionate to the advertised use case.
  • No training of competing models on generated Output.

5A.4. What we treat as abuse indicators

When evaluating load we look not at an absolute number (which doesn't exist for unlimited plans) but at behavioural patterns:

  • generation volume anomalously above the average of comparable users on the same plan (cohort-relative);
  • bot-like request patterns (uniform intervals, absence of human-driven pauses, thousands of concurrent sessions on one Account);
  • requests inconsistent with the use case declared at onboarding;
  • multi-tenant indicators — wide range of source IPs, input languages, content topics not consistent with a single-end-user pattern.

When such indicators appear we may apply the enforcement measures in Section 7 of this Policy based on abuse of the sharing rules (Section 5A.3) or Section 1 — not based on a volume threshold.

6. Reporting abuse

If you become aware of any violation of this AUP — including abuse of your own voice by another user — report it to:

abuse@velluto.io

Include: a description of the issue, URLs or Account identifiers if known, evidence (screenshots, audio samples, links), your contact details, and (if relevant) proof that you are the affected party.

We aim to acknowledge reports within 72 hours and to resolve them within 14 days. We may need additional information for verification.

7. Enforcement

7.1. We may, in our sole discretion and without notice:

  • remove or refuse to process Content;
  • disable or delete Voice Models;
  • restrict features available to your Account;
  • suspend or terminate your Account;
  • forfeit prepaid fees as liquidated damages for breach;
  • cooperate with law-enforcement investigations and produce records under valid legal process;
  • publicly disclose the fact of enforcement action in cases of large-scale or public-interest violations (without disclosing personal data of the violator beyond what is legally required).

7.2. For a first minor violation we may issue a warning and request remediation before stronger action.

7.3. For violations of Sections 2.1, 2.2, 3.1, 3.2, 3.5, or any criminal conduct, enforcement is immediate, irreversible, and without warning.

7.4. Enforcement action under this AUP does not waive any other right we have under the Terms or applicable law.

8. No safe harbour for circumvention

Attempting to circumvent the AUP — through prompt manipulation, multi-account abuse, sock-puppet accounts, fragmenting prohibited content across generations, or laundering output through other services — is itself a violation of the AUP and a material breach of the Terms.

9. Changes

We may update the AUP from time to time to address new abuse patterns and new legal requirements. Material changes will be notified with at least 30 days' notice; urgent changes addressing severe abuse may take immediate effect with notice given as soon as practicable.

10. Contact

Questions about this document? legal@velluto.io.