Trust & Safety

1. Our position

Lumino is an AI-powered creative suite operated by Lumino Labs LLC, a Delaware-registered company. Our tools generate content that can be persuasive, lifelike, and easily shared. We take the safety implications of that seriously. This page describes what we do not allow, how we enforce it, and how to report a problem.

2. Prohibited uses

You may not use Lumino, in whole or in part, to:

• Impersonate any real person, public figure, or organization without their explicit, documented consent.
• Create deepfakes or other synthetic media intended to deceive, defraud, defame, harass, or sexually exploit any person.
• Generate non-consensual intimate imagery, child sexual abuse material, or any content sexualizing minors. Such content is prohibited unconditionally; no creative framing makes it acceptable.
• Clone a voice you do not own without the speaker's explicit, documented permission.
• Produce fraudulent documents (invoices, receipts, flight tickets, bank statements, contracts) for use in real financial, legal, tax, immigration, insurance, expense, or visa processes.
• Operate a fake charity, fake bank, fake courier, fake brokerage, fake celebrity profile, or fake virtual identity intended to deceive third parties.
• Use virtual phone number rentals to bypass platform security controls, harass recipients, or commit fraud.
• Generate content promoting violence, hate, discrimination, terrorism, illegal goods, or self-harm.
• Infringe intellectual property, including trademarks, logos, and likenesses you do not have rights to use.
• Circumvent, disable, or remove watermarks, content-provenance markers, or AI-generated labels applied by Lumino or by upstream model providers.

3. Impersonation policy

Impersonation is the prohibition we enforce most strictly because it underlies most of the harms our tools could enable.

• You may not use any Lumino feature to depict, voice, or imitate a real person (living or recently deceased) without their explicit, documented consent.
• This applies to Avatar Video, Voice Clone, Photo Studio, Live Transform, AI Writer, and Document Generator outputs alike.
• Lumino does not ship Live Transform style presets that target named real people. Reference images supplied for Live Transform must be of yourself or a person who has consented in writing.
• Public figures, brands, and trademarks are not fair game by default; their names and likenesses are protected under impersonation, defamation, publicity-rights, and trademark law in most jurisdictions, and our policy reflects that.

4. Voice cloning consent

Voice Clone Studio creates a model of a human voice from a short audio sample. Voice cloning without consent is a serious harm vector and we treat it that way.

• By submitting an audio sample, you affirm that the voice in the sample is either yours or that you have explicit, documented permission from the speaker to clone it.
• Cloning the voice of a public figure or any third party without consent is prohibited and will result in immediate account termination, deletion of the voice clone, and forfeiture of credits.
• Cloned voices remain tied to the account that created them. They cannot be transferred between accounts.
• You can delete a cloned voice at any time from Voice Clone Studio. Deletion is irreversible.

5. Live Transform safety

Live Transform restyles a creator's own live camera feed in real time. We apply additional controls to this surface because real-time face transformation is one of the most sensitive AI categories.

• Each session begins with an explicit on-screen consent affirming that you are the person on camera or you have the subject's documented consent, and that you will not use the output to impersonate or deceive.
• Built-in presets are limited to artistic styles (Cyberpunk, Anime, Renaissance, Cartoon and similar). We do not ship presets that target named real people.
• Custom reference images uploaded by you must depict yourself or a person who has consented in writing. Uploading the face of a real third party without consent is prohibited.
• Live Transform output is logged at the session level (account ID, timestamp, consent version, model used). We use these logs only for abuse review.
• Where the upstream model applies a content-provenance watermark, you may not attempt to remove or obscure it. Where we offer unwatermarked output to an approved account, we still mark the content as AI-generated through other means (metadata, in-product labels, or post-session disclosure).

6. Simulated tools (Banking, Broker, Courier, Charity, Celebrity, Documents, Virtual Numbers)

Several Lumino tools generate simulated versions of real-world surfaces: bank portals, brokerage dashboards, courier tracking pages, fundraiser pages, celebrity profile pages, and document templates such as invoices, receipts, and flight tickets. These exist for creative, demonstration, prototyping, and educational use.

• Nothing produced by these tools is real. They do not move money, settle securities, ship parcels, accept donations, verify identities, or constitute valid financial, travel, legal, or commercial documents.
• Outputs from these tools may not be submitted as evidence in real financial, legal, tax, immigration, insurance, audit, expense-reimbursement, or visa processes. Doing so is fraud and is grounds for termination and reporting to authorities.
• You may not use these tools to impersonate a real bank, broker, exchange, courier, charity, public figure, business, or telecommunications provider.
• You may not use virtual phone numbers to bypass another platform's security or terms of service.
• These prohibitions are enforced through the same enforcement framework as everything else: account termination, credit forfeiture, and reporting where appropriate.

7. Account-tied access & monitoring

All Lumino features are gated behind authenticated user accounts. There is no anonymous public generation endpoint.

• Every generation, voice clone, simulated account, Live Transform session, and document is logged against the originating user account.
• Rate limits are enforced per account (for example, 5 video generations per hour and 20 per day) to limit large-scale abuse.
• Activity is reviewable in our administrative tooling. We can identify, suspend, and act on accounts associated with abuse.
• We retain abuse-relevant logs for a reasonable period to respond to complaints and law-enforcement requests.

8. Content provenance

We support clear labeling of AI-generated content. Where an upstream model (Decart, fal.ai, HeyGen, ElevenLabs, Google) applies a watermark or provenance marker to the output, that marker remains in place by default.

• You may not strip, obscure, or alter provenance markers applied by us or by an upstream model.
• Where we operate a removed-watermark agreement with an upstream provider, we mark content as AI-generated through other means (metadata, in-product disclosure, or post-session notice to the user).
• We support C2PA-style provenance metadata where the upstream pipeline emits it.

9. Enforcement

We enforce this policy through automated controls and human review.

• First-line controls include rate limits, file-type validation, upstream model moderation filters, and content-policy checks on user-supplied text.
• Reported violations are reviewed by a human. Confirmed violations result in content removal, account suspension or permanent termination, credit forfeiture, and where appropriate, reporting to law enforcement and to affected platforms or rights-holders.
• Severe violations (CSAM, non-consensual intimate imagery, credible threats of violence) are reported to NCMEC and to relevant authorities without warning and without refund.

10. Reporting abuse

To report content or accounts that violate this policy, email abuse@lumino.work (forwards to support). Include the URL or content reference, a brief description of the violation, and your contact details if you would like a follow-up. We aim to respond within 3 business days. For takedown requests involving your own likeness, voice, or copyrighted material, please include verifiable proof of identity or ownership; we will act on credible, properly-supported requests.

11. Working with upstream providers

Lumino integrates with third-party AI providers (Decart, fal.ai, HeyGen, ElevenLabs, Google AI). Where a provider imposes additional rules on what their model may be used for, those rules apply on top of this policy. We share abuse signals with upstream providers where there is a credible safety reason to do so.

12. Changes to this policy

We update this policy as the product and the regulatory landscape evolve. Material changes are noted by the "Last updated" date at the top. Continued use of Lumino after a change constitutes acceptance of the revised policy.

13. Contact

For Trust & Safety questions, email abuse@lumino.work or support@lumino.work.

Operated by Lumino Labs LLC, registered in Delaware, United States.