Usage Policy

The short version

Kaldune is a studio for real work. There's a lot it will take on. There's a short list it won't, and that list isn't ours alone. Because the work in your room is done by AI from several third-party providers, we follow their rules, and we add our own on top. When they differ, the stricter one applies.

We don't promise to take on anything. If a brief asks for something on the list below, the room won't do it, and we may pause or close an account that keeps asking.

Here's the plain version. The exact standards follow.

• Nothing illegal, and nothing built to harm people. No weapons, no attacks on computer systems or critical infrastructure, no content that exploits or endangers children, no fraud or deception, no harassment.
• Nothing that violates someone's rights. Not their privacy, not their identity, not their voice or face, not their intellectual property.
• Be honest about AI. For high-stakes uses, anything touching someone's legal, medical, financial, or employment situation, a qualified human has to review the work, and the people affected should know AI was involved.
• Don't misuse the studio itself. Don't try to extract the underlying model, train a competitor on it, or resell it.
• These rules apply wherever you are. Kaldune is offered only in the countries and regions our AI partners support, and never where U.S. sanctions apply. This policy applies to every user in every country where it is offered, and so does the law where you are: you're responsible for using the studio lawfully in your own location, and where local law is stricter than this policy, local law also applies.

Last updated: 30 October 2026 · Version 1.8

Exact policy

This Usage Policy ("Policy") governs your use of the Kaldune service operated by Studio58, Inc. ("Kaldune," "we," "us," or "our"), a Delaware corporation. It supplements the Terms of Service. If you violate this Policy, we may block or modify the work, and throttle, suspend, or terminate your access, as described in the Terms.

1. How this Policy relates to the AI providers' policies

1.1. To make Kaldune's work, the studio calls on AI models from several third-party providers (for language, images, video, voice, sound and music), and on open models we run on our own cloud infrastructure. Your use of Kaldune is subject to the usage policies of the providers whose models the studio calls on for your work, which we pass through to you. The providers are listed on our sub-processors page.

1.2. This Policy is at least as restrictive as the strictest of those providers' usage policies and incorporates them by reference. Any use a provider prohibits is prohibited on Kaldune, whether or not it is separately named here. Where this Policy and a provider's policy differ, the more restrictive rule applies.

1.3. Because we rely on the providers' published policies, we keep this Policy current against them, and we review it whenever a provider changes its policy. Each provider publishes its own policy on its own website.

1.4. This Policy applies globally. The Service is offered only where Section 1.5 says (see also Terms of Service Sections 13 and 14.4). This Policy applies to every user regardless of location. It does not replace the law that applies to you where you are: you are responsible for complying with the laws of your own jurisdiction, and where those laws prohibit conduct this Policy permits, or impose stricter limits, the stricter rule applies to your use.

1.5. Where the Service is available. Kaldune is offered only in the countries and regions our AI partners support, and never where U.S. sanctions apply. It is not offered in Afghanistan, Antarctica, Belarus, China (including Hong Kong and Macao), Cuba, Iran, Kosovo, Myanmar, North Korea, Russia, Syria, Venezuela, Western Sahara, Yemen, or the Crimea, Donetsk, Kherson, Luhansk or Zaporizhzhia regions of Ukraine. Access from every country and territory on this list is blocked at the network layer, and payments from them are refused; the listed regions of Ukraine are prohibited by the Terms of Service Section 13. If you travel to one of those places briefly, do not sign in to the Service until you leave.

2. Universal Usage Standards: uses that are not allowed

You may not use Kaldune, and you may not direct the room, to do any of the following. These standards reflect the usage policies of the AI providers whose models the studio calls on, and apply to every user, consumer and business alike.

2.1. Violate the law or engage in illegal activity. Do not use the Service to plan, carry out, or facilitate unlawful conduct.

2.2. Compromise critical infrastructure. Do not use the Service to damage, disrupt, or gain unauthorized access to critical infrastructure: power, water, transportation, financial, communications, or similar systems.

2.3. Compromise computer or network systems. Do not use the Service to create or distribute malware, gain unauthorized access to systems or data, conduct denial-of-service attacks, engage in spoofing or social engineering, or otherwise attack the security, integrity, or availability of any system. (Defensive security work done with the system owner's consent is not prohibited, and good-faith security research on the Service itself, conducted within the safe harbor in Terms of Service Section 19, is treated as authorized; report vulnerabilities to security@kaldune.ai.)

2.4. Develop or design weapons. Do not use the Service to develop, design, produce, or facilitate weapons, explosives, or dangerous materials, including chemical, biological, radiological, nuclear, or high-yield explosive weapons.

2.5. Incite violence or hateful behavior. Do not use the Service to promote, incite, or facilitate violence, or to produce content that harasses, demeans, or promotes hatred against people based on a protected characteristic.

2.6. Compromise privacy, identity, or intellectual-property rights. Do not use the Service to violate a person's privacy rights, to misuse confidential or personal information, to identify, track, profile, or surveil a person in violation of their rights or applicable law, or to infringe a third party's intellectual-property rights (including copyright, trademark, or trade secret). We respond to copyright complaints and maintain a repeat-infringer policy under the DMCA process in Terms of Service Section 18; repeated infringement can result in termination.

2.7. Compromise children's safety. Do not use the Service to create, distribute, or facilitate content that sexually exploits or abuses a minor, or to endanger a child in any way. A minor is anyone under 18, regardless of jurisdiction. We report child sexual abuse material and the coercion or enticement of a minor to the relevant authorities, as do the AI providers.

2.8. Create psychologically or emotionally harmful content. Do not use the Service to produce content designed to manipulate, coerce, or cause serious psychological or emotional harm, including content that promotes self-harm, disordered eating, or suicide.

2.9. Create or spread misinformation. Do not use the Service to knowingly create or spread false information intended to deceive, including about health, safety, or civic matters.

2.10. Undermine democratic processes or run targeted campaign activities. Do not use the Service to interfere with elections or democratic processes, or to run prohibited political-campaign or targeted-influence activities, in violation of a provider's policy or applicable law.

2.11. Misuse for criminal justice, censorship, surveillance, or prohibited law-enforcement purposes. Do not use the Service for social scoring, mass surveillance, censorship, or law-enforcement uses a provider's policy prohibits. Nor for recognizing people's emotions outside medical or safety uses, sorting people by biometric data into categories such as race, religion, sexual orientation or political views, publishing someone's private details to expose or target them (doxxing), or tracking people without their consent.

2.12. Engage in fraudulent, abusive, or predatory practices. Do not use the Service for scams, phishing, spam, plagiarism-for-hire, academic dishonesty sold as a service, deceptive commercial practices, or predatory lending or similar schemes. Nor for deceptive campaigns or fake activity: fake reviews, astroturfing, bots posing as people, sponsorship you hide, or manipulating search results or AI sources. Do not hide from the people who interact with the work that AI was used to make it.

2.13. Abuse the platform. Do not use the Service to circumvent our safeguards, resell or share access in violation of the Terms, interfere with or degrade the Service for others, or abuse the billing system, including fraudulent chargebacks, redeeming or trafficking promotional credits in violation of their terms, or exploiting the credits, subscription, or refund process (see Terms of Service Sections 7.4 to 7.7). This includes running multiple accounts or automated sign-ups to get around limits or bans, jailbreaking the studio or attacking it with prompt injection, and reselling or proxying access to it.

2.14. Generate sexually explicit content. Never anything sexual involving a minor.

2.15. Model misuse. Do not use the Service or its Outputs to (a) build or train a competing AI model, (b) reverse-engineer, extract, or duplicate the underlying models ("model scraping"), or (c) resell the underlying model capability, except as we expressly permit in writing.

2.16. Abuse a published page. Pages you publish through the Service (see Terms of Service Section 3.9) live on our domain, so a published page that violates this Policy, or that hosts fraud, phishing, impersonation, or malware, is our problem the moment it is up. We may take a published page down immediately, without prior notice, and repeated abuse may end your right to publish or your account. A takedown removes the public page, not the underlying work in your studio. The same rules apply to apps published at their own address on kaldune.app.

2.17. Clone a voice or fake a real person. Do not use the Service to clone or re-voice a real, identifiable person's voice, to make an image, video or avatar of a real, identifiable person, or to lip-sync words onto a real person's face or voice, without that person's consent (or, for a minor, a parent's or guardian's). Do not use a cloned voice, a likeness or lip-sync to impersonate anyone, and do not present synthetic audio or video of a real person as a genuine recording. Do not ask the studio to imitate a specific living singer's or speaker's voice so that listeners would take it for theirs. Some providers refuse a real person's likeness outright, and where they do, the room won't make it.

2.18. Misuse music and lyrics. Do not use the Service to reproduce the lyrics or melody of a song you do not have the rights to, or to pass off a piece as the work of a real artist. Lyrics you give the studio to sing must be yours or licensed to you. Lyrics that harass, defame or incite, or that break any other rule in this Section, are not allowed in a sung piece any more than in writing.

2.19. Collect personal data through an app without telling people. An app you publish through the Service that collects personal data from its visitors must tell them what it collects and why, and must not collect special-category data (health, biometrics, sexual life, religious or political views) or payment card numbers.

3. High-risk use requirements

3.1. Some uses carry an elevated risk to people because they bear on decisions vital to a person's rights, safety, livelihood, or health. These "high-risk uses" include work that materially informs legal, medical, financial, credit, insurance, employment, housing, education-access, healthcare-access, public-benefits, immigration or legal-status, or similar consequential decisions about identifiable people, and AI that controls physical equipment.

3.2. For high-risk uses, a qualified person reviews the work before it is used, and the people affected are told AI was involved. In line with the AI providers' requirements, you must:

• (a) Keep a qualified human in the loop. A person with relevant expertise must review the work before it is relied on or acted on.
• (b) Disclose AI involvement. You must tell the people affected that AI was involved in producing the material, and, where you deploy Kaldune-produced material in an automated service that interacts directly with your own users, disclose that they are interacting with an AI system rather than a human.

3.3. These requirements are in addition to the professional-advice disclaimer in the Terms (Section 9.3): Kaldune does not provide legal, medical, financial, or other professional advice, and its Outputs are not a substitute for a qualified professional.

4. What the studio will take on

We say plainly what we won't do; we do not promise to do everything else. Whether the room can take on a given brief depends on the brief, the Service's current capability, this Policy, and the AI providers' policies. If a brief sits near the edge of this Policy and you're not sure, ask us at support@kaldune.ai before you rely on it.

5. Enforcement

5.1. We use automated and, where necessary, human review to enforce this Policy. Where an AI provider's own safeguards block or modify an Output, that result flows through to your room.

5.2. If we believe you have violated this Policy, we may block or modify the work, throttle, suspend, or terminate access, and, where the law requires or permits and a serious harm is involved (for example, child safety), report to the relevant authorities. Suspension and termination follow the Terms.

6. Changes

We update this Policy as our Service, the AI providers' policies, the law, and the risks evolve. When a provider changes its policy, we review and update this Policy so it stays at least as restrictive. The "Last updated" date and version at the top reflect the current version.