Acceptable Use Policy

Version 0.1 (draft) · Effective: {{EFFECTIVE_DATE}} · Last updated: 24 October 2026

In short (plain language)

DamruVani makes realistic voices. That power can hurt people if it is misused. So:

• Never pretend to be someone else with DamruVani — no cloning or imitating real people without their consent, and never famous people or politicians.
• Never use it to cheat people — no fake bank, KYC, OTP, police, "digital arrest" or customer-care calls.
• No political deepfakes and no election content that misrepresents anyone.
• Nothing sexual, and nothing involving children in a harmful way.
• No hate, threats, harassment or fake news.
• Label AI audio honestly and never remove our labels.

If you break these rules we may block the content, disable voices, suspend or close your account, and report serious cases to the authorities. This policy is part of our Terms of Service.

1. Impersonation and consent

You must not:

1.1 Clone, imitate or generate the voice of any real person without their consent — including through a "soundalike" description or a recording taken from a video, call or social media. Voice Lab clones are allowed only for your own voice (Voice Clone Consent).

1.2 Imitate a well-known person — politicians, government officials, film and TV stars, singers, sportspeople, journalists, business leaders, religious leaders or other public figures — even with a disclaimer and even as parody. We keep a blocklist of names (in English and Indian scripts) and reject matching voice names, descriptions and requests. [LAWYER: parody / satire carve-out]

1.3 Imitate a person who has died. [LAWYER: post-mortem personality rights]

1.4 Speak as a deity, guru, saint or religious leader in the first person ("I am Shiva / Krishna / Guru … speaking"). Devotional narration, bhajans and reading scriptures in a narrator's voice are fine.

1.5 Present AI audio as a genuine recording of a real person or a real event in a way that is likely to deceive.

1.6 Use a DamruVani Library Voice to suggest that the real voice actor said something, or try to copy or extract a Library Voice.

2. Fraud, scams and deception

You must not use DamruVani for:

2.1 Scam or phishing calls and messages — for example fake bank, UPI, KYC, OTP, "your account will be blocked", electricity-disconnection, courier, lottery, job or investment calls; fake police, CBI, customs or court officials ("digital arrest"); fake customer-care numbers; or extortion.

2.2 Content asking people to share OTPs, PINs, passwords or card details, to click a link, to install an app or to transfer money under false pretences.

2.3 False documents or false electronic records, fake evidence, or fake official announcements (from government, police, courts, the RBI, banks, hospitals or emergency services).

2.4 Bypassing voice-based security or identity checks (for example "voice password" systems at banks or call centres).

Requests that use a cloned voice, and bulk or template requests, are checked strictly for scam patterns. If a check is uncertain, the request is blocked.

3. Elections and political content

You must not:

3.1 Generate the voice or name of any real politician, candidate or party functionary.

3.2 Create political deepfakes, or election-related content that misrepresents any person's identity, voice, conduct or statements, or misleads voters about how, when or where to vote.

3.3 Use cloned voices for political advertisements, campaign robocalls or mass political messaging. During election periods (when the Election Commission of India's Model Code of Conduct is in force) we may block all political content made with cloned voices and send it for manual review.

3.4 Any lawful political content you make must follow the Election Commission of India's advisories on labelling AI-generated campaign content (for audio, the label must be announced in the first 10% of the clip, as reported for the October 2025 advisory) [VERIFY: current ECI advisory text], and the Representation of the People Act, 1951.

4. Children

4.1 The Service is only for people 18 or older.

4.2 You must not clone a child's voice or upload a child's voice for cloning.

4.3 Absolutely no child sexual abuse or exploitation material, in any form. We report it to the authorities as the law requires (including under the POCSO Act, 2012).

4.4 Child-sounding Library Voices (performed by adults) must never be used for sexual, romantic or violent content, to imitate a real child, to ask children for personal information, or for scam or prank calls. All requests using these voices go through strict moderation. Advertising aimed at children using these voices needs our prior written approval.

5. Sexual content and intimate imagery

5.1 No sexually explicit, obscene, pornographic or vulgar content.

5.2 No non-consensual intimate content of any person, including AI-generated content, and nothing that invades a person's bodily privacy.

6. Hate, harassment and violence

You must not create content that:

6.1 Promotes hatred, discrimination or violence against people based on religion, caste, tribe, race, ethnicity, region, language, gender, sexual orientation, disability or similar characteristics — in any language or dialect.

6.2 Threatens, harasses, bullies, intimidates, doxxes or defames anyone.

6.3 Promotes terrorism, incites violence or endangers public order, the sovereignty and integrity of India, or friendly relations with foreign states.

6.4 Encourages suicide or self-harm.

7. Illegal and dangerous activities

You must not use the Service to:

7.1 Help make or obtain explosives, weapons or ammunition, or illegal drugs.

7.2 Promote illegal gambling, betting or online money games. [VERIFY: Promotion and Regulation of Online Gaming Act, 2025]

7.3 Make automated or bulk calls to people who have not agreed to receive them, or commercial calls that break telecom rules. [VERIFY: TRAI Telecom Commercial Communications Customer Preference Regulations, 2018, as amended]

7.4 Break any law in force in India.

8. Intellectual property and others' rights

8.1 Only convert text, and only upload audio or video, that you have the right to use. For example, do not create audiobooks of copyrighted books, or dub other people's videos, without permission.

8.2 Do not violate anyone's privacy, publicity or personality rights.

9. Misinformation

Do not create or spread content you know (or should reasonably know) is false and likely to cause harm — for example fake news bulletins, fake emergency alerts, fake disaster warnings or false health claims.

10. Honest labelling

10.1 Do not remove, alter, hide or suppress DamruVani's AI labels, metadata, watermarks or other provenance markers, or use tools to do so.

10.2 When you publish AI audio, disclose that it is AI-generated wherever the law or the platform requires it. See AI Content Disclosure.

11. Commercial use during the beta

During the public beta, you may use DamruVani and its outputs only for non-commercial purposes (Terms §4).

12. Protecting the Service

You must not:

12.1 Scrape the website, voice catalogue, previews or API, or download previews in bulk.

12.2 Get around limits, queues or free-credit rules — for example with multiple accounts, shared or resold accounts, or automated sign-ups.

12.3 Reverse-engineer, decompile or try to extract our models, voices or source code, except where the law expressly allows it.

12.4 Use the Service or its outputs to build, train or benchmark-for-training a competing speech model or dataset.

12.5 Probe, scan, load-test or attack the Service, or upload malware or files designed to harm our systems, without our written permission. To report a security issue, write to security@{{DOMAIN}}. [OWNER: create mailbox or route to grievance@]

12.6 Share your API keys or put them in public code.

13. How we enforce this policy

13.1 Prevention. We use automated tools — including AI text moderation in every supported language, a blocklist of well-known names, scam-pattern checks and voice-clone verification — to stop prohibited content before it is generated. Where our moderation is less certain (for example in beta languages), requests that use cloned voices are blocked when in doubt.

13.2 Review. People on our team review reports, flagged content and some voice clones. Reviewers can access content only when it is linked to a report or a safety flag, and every access is logged.

13.3 Actions. Depending on how serious the violation is, we may: refuse the request; warn you; remove content or disable share links; disable a voice or an API key; suspend or terminate your account (without destroying evidence); report the matter to the appropriate authority where the law requires; and, in accordance with applicable law, disclose the violating user's identity to a complainant who is a victim.

13.4 Your liability. Misusing AI tools can lead to penalties or punishment under the law, including the Information Technology Act, 2000 and the Bharatiya Nyaya Sanhita, 2023. See the notice in Terms §12.

13.5 Appeals. If you think we made a mistake, write to our Grievance Officer (/legal/grievance). Where the IT Rules apply, you may also appeal to the Grievance Appellate Committee.

14. Reporting misuse

Use the Report button on any audio, voice or share page, or write to grievance@{{DOMAIN}}. If your own voice or identity has been misused, say so — we treat it as urgent. Timelines: /legal/grievance. Victims of online fraud should also report at cybercrime.gov.in or call 1930. [VERIFY: helpline number and portal current]

15. Third-party model licence terms

Model 1 (beta) includes components built with Higgs Materials (Boson AI) and Meta Llama 3 (Credits). Their licences require every use of the Service to follow the Meta Llama 3 Acceptable Use Policy (https://llama.meta.com/llama3/use-policy), which forms part of this policy. Most of it is already covered above; in addition, you must not use the Service or its outputs to:

15.1 Practise a regulated profession (for example legal, medical or financial advice) without the required authorisation, or present outputs as such advice.

15.2 Generate content for military, warfare, nuclear or espionage purposes, or for operating critical infrastructure, transport systems or heavy machinery.

15.3 Promote or encourage suicide, self-harm or eating disorders.

15.4 Collect, process or expose sensitive personal information (for example health or demographic data) of others without the rights and consents the law requires.

15.5 Create spam or false online engagement (for example fake reviews, fake testimonials or bot comments), or present outputs as human-generated.

15.6 Discriminate against people in employment, credit, housing or other essential goods and services.

If the Meta Llama 3 Acceptable Use Policy and this policy differ, the stricter rule applies. [LAWYER: incorporation by reference vs restating; Higgs licence §1.b.iv — README L12]