Content Removal & NCII / Deepfake Takedown Policy

This Content Removal & NCII / Deepfake Takedown Policy explains how any person depicted in intimate imagery (or their authorized representative) can ask Velum AI operator [to be configured] ("we", "us", "the operator"), operator of https://velum.ai (the "Service"), to remove that imagery, and how quickly we act. It implements the U.S. TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks Act, 2025) and reflects our obligations under card-network rules, applicable privacy law, and our own zero-tolerance standards.

This policy applies to all content generated, stored, hosted, delivered, or otherwise processed through the Service, including AI-generated ("synthetic") photos and videos, training images, shared folders, and any material a creator uploads or that we deliver.

This is a removal channel for non-consensual intimate imagery and sexual deepfakes of real people. Copyright complaints follow the separate DMCA process in Section 12; requests to access or delete your personal data follow the privacy-rights process, also described in Section 12.

"Intimate visual depiction" or "NCII" (non-consensual intimate imagery) means a photograph, video, or digital image that shows a real, identifiable person nude, with exposed genitals, anus, or (for a person who identifies as female) post-pubescent nipple/areola, or engaged in sexually explicit or simulated sexual conduct, where that person did not consent to its creation, generation, or publication, or where publication was reasonably expected to remain private.

"Digital forgery" or "deepfake" means intimate visual content created or altered by artificial intelligence, machine learning, or other technological means so that it appears to authentically depict a real, identifiable person, whether or not the content is labeled as synthetic.

"Depicted individual" means the identifiable real person shown, or realistically appearing to be shown, in the content. "Authorized representative" means a parent or legal guardian of the depicted individual, or an attorney, agent, or other person with written authority to act on the depicted individual's behalf.

"Covered content" means any NCII or digital forgery of a real person, whether it appears in a creator's training set, in generated output, in a shared folder, in delivered files, in backups, or anywhere else on the Service.

The Service is built on a single, non-negotiable rule: a creator may only request and receive an AI likeness model ("clone" / LoRA) of their own verified likeness, and only after they have proven their identity and age and signed explicit consent. We never knowingly create, train on, or generate the likeness of anyone other than the consenting, verified account holder.

Because of this rule, non-consensual intimate imagery of a third party should never exist on the Service. If such content is present, it is by definition a violation of our Terms of Service and of this policy, and we will remove it and act against the responsible account.

This policy is the safety net behind that rule: it gives any depicted person a fast, direct way to have their imagery removed even if someone attempted to bypass our verification and consent controls.

A removal request may be submitted by any of the following: the depicted individual themselves; a parent or legal guardian if the depicted individual is a minor, deceased, or otherwise unable to act; or an authorized representative acting with written permission.

You do not need to be a user of the Service, and you do not need to identify the account or creator responsible. You only need to identify the content and confirm that it depicts you (or the person you represent) without consent.

If the depicted individual is or may be a minor, do not rely on this form as the primary channel: such material is treated as suspected child sexual abuse material (CSAM), removed on discovery, reported to the National Center for Missing & Exploited Children (NCMEC), and handled under Section 12. You may still notify us at the contact in Section 15 so we can act immediately.

Submit a removal request through our dedicated reporting channel: «[TO BE SET: URL of the online NCII/deepfake takedown form]» or by email to «[TO BE SET: dedicated takedown email address, e.g. [dedicated email address to be configured]]». Requests may also be sent to our general support address, Support contact pending configuration, and will be routed to the takedown team.

So we can act quickly, please include: (a) enough information to locate the content (a direct link, screenshot, file name, delivery reference, or creator/stage name); (b) a statement that the content depicts you, or the person you represent, in an intimate way; (c) a statement that you (or the depicted person) did not consent to its creation, generation, or publication; (d) your name and a means of contact; and (e) if you are a representative, a brief statement of your authority to act.

Include an electronic signature (typing your full legal name is sufficient) and the statement: "I have a good-faith belief that the identified content is a non-consensual intimate image or digital forgery of the depicted person, and the information in this request is accurate." We do not require notarization, a police report, a lawyer, or any fee, and we accept requests in Spanish and English.

When we receive a valid removal request, we will remove or disable access to the identified content as soon as possible and no later than 48 hours after receipt, consistent with the TAKE IT DOWN Act. The 48-hour clock starts when a request containing the information in Section 5 reaches our reporting channel.

If a request is incomplete or we cannot locate the content, we will promptly tell you what is missing; the 48-hour period runs from receipt of the completed request. We will not use minor formatting issues to delay a good-faith request from a depicted person.

Removal means the content is taken down from public or shared availability and from delivery to any account. Where content is unlawful (for example, CSAM), we will also preserve the records necessary for law enforcement before deletion, as described in Section 12.

After removing reported content, we will make reasonable efforts to identify and remove any known identical or substantially similar copies across the Service, including copies in the responsible account's generated output, training sets, shared folders, and backups scheduled for the next routine purge, as required by the TAKE IT DOWN Act.

Where technically feasible we use content fingerprinting/hashing to detect and block re-uploads or re-generation of the same imagery. No automated system is perfect; if you find a copy we missed, tell us and we will remove it under the same 48-hour standard.

We may retain a non-visual hash or minimal identifier of removed NCII solely to prevent the same material from reappearing, and for no other purpose.

If you are the depicted person and you already hold a verified account with us, you may report content of yourself directly from your account; because your identity and likeness are already verified, we prioritize these reports and typically act well within the 48-hour window.

If you are the depicted person but do not have an account, you may ask us to confirm it is you using a light-touch check (for example, a selfie matching the depicted likeness, or verification of control of a contact channel). We will request only the minimum needed and will never demand more documentation than necessary to act; where there is any doubt, we resolve it in favor of the depicted person and err toward removal.

Nothing in this expedited path delays the 48-hour deadline: if a request is facially valid, we act on it and resolve any identity questions in parallel.

To guard against abuse and against removing content that is in fact consensual and lawful (for example, a verified creator's own authorized content being targeted by an impostor), we may take reasonable steps to confirm that the requester is the depicted person or a duly authorized representative.

For representatives, we may ask for written authorization or proof of legal authority (such as guardianship). For the depicted person's own request, we keep verification minimal per Section 8. We will not disclose the requester's identity or the contents of the request to the affected creator except as needed to act, to comply with law, or to establish or defend a legal claim.

We will confirm to the requester when the content has been removed.

We notify the account responsible for the content that material was removed under this policy, to the extent that doing so does not endanger the depicted person or interfere with a law-enforcement matter. We may suspend or terminate any account that uploaded, trained on, generated, or shared NCII or a deepfake of a person other than the verified account holder.

Because our only-your-own-likeness rule forbids third-party likenesses, a confirmed non-consensual depiction of someone other than the account holder is treated as a serious violation and generally results in immediate termination, in addition to removal.

Where content is or may be illegal, we preserve the minimum necessary evidence and report to the appropriate authorities (including NCMEC for suspected CSAM) as described in Section 12.

This channel exists to protect real people. Knowingly submitting a materially false removal request — for example, falsely claiming to be the depicted person, falsely claiming lack of consent in order to remove a creator's own lawful content, or filing requests to harass a creator — is prohibited, may violate applicable law, and may be reported to authorities.

We may decline or reverse action on requests we determine in good faith to be fraudulent or abusive, and we may act against accounts that misuse this process. Even so, we will continue to err on the side of protecting a genuine depicted person; abuse controls are never used to delay a good-faith takedown.

Submitting a request does not waive any of your legal rights, and using this process is not a precondition to pursuing other remedies.

Copyright (DMCA): If your complaint is that content infringes your copyright, rather than that it depicts you intimately without consent, use our DMCA notice-and-takedown process; send notices to our designated agent «[TO BE SET: DMCA designated agent name, physical US address, and email as registered with the U.S. Copyright Office]». NCII removal under this policy does not require any copyright claim.

CSAM and minors: We have zero tolerance for any content that involves or appears to involve a minor. Such material is removed on discovery, reported to NCMEC and, where applicable, to law enforcement, and the account is terminated, consistent with our 18 U.S.C. § 2257 statement and our Terms.

Privacy rights (GDPR/CCPA and similar): You may separately request access to or deletion of your personal data. EU/UK/EEA data-subject and NCII matters may also be directed to our representatives: «[TO BE SET: EU Art. 27 GDPR representative name and address]» and «[TO BE SET: UK GDPR representative name and address]». Removal under this policy and erasure of underlying biometric/training data are coordinated so that removed NCII is not retained beyond what Section 7 permits.

Producer status and provenance: As the producer of the AI-generated adult content on the Service (including for 18 U.S.C. § 2257 purposes), we maintain age and consent records for the verified creator behind each clone and apply AI-provenance labeling to generated output; none of this reduces a depicted person's right to removal under this policy.

Card networks: We maintain this removal process in part to meet Visa and Mastercard requirements for adult-content platforms, including the prompt removal of non-consensual content.

If we removed content that was in fact your own verified, consensual, lawful content and it was taken down in error, you may ask us to review and restore it by contacting us at the address in Section 15. We will review promptly and restore content that we confirm is lawful and consensual.

We will not restore any content that a depicted person has validly reported as non-consensual, any suspected CSAM, or any content whose restoration would violate the law or card-network rules.

We keep records of removal requests and the actions we took (what was reported, when it was received, when it was removed, and any duplicate-removal efforts) for as long as needed to comply with the TAKE IT DOWN Act, card-network rules, and other applicable law, and to defend against or pursue legal claims.

We treat removal requests and requester identities as confidential and disclose them only to act on the request, to comply with law, or to establish or defend legal claims. Removed NCII is not retained except as a non-visual hash under Section 7 or as evidence preserved for law enforcement under Section 12.

NCII / deepfake takedown requests: «[TO BE SET: URL of the online takedown form]» or «[TO BE SET: dedicated takedown email address]». General support: Support contact pending configuration.

Designated point of contact and custodian for removal and 2257 records: «[TO BE SET: name of designated agent / custodian of records]», «[TO BE SET: US street address for the custodian of records]». Operator: Velum AI operator [to be configured], https://velum.ai.

We may update this policy to reflect changes in the law, card-network rules, or our operations. The current version is always posted at https://velum.ai. Material changes take effect when posted; continued use of the Service after posting constitutes acceptance of the updated policy.