Acceptable Use Policy

RiffAds is an AI video ad platform run by [Company legal name] ("we", "us"). You use it to make ads with AI actors, voices, scripts, images and video, in the web app at app.riffads.com and through our API, MCP server, CLI and Agent Skills.

This Acceptable Use Policy explains what you must not do with RiffAds. It is part of our Terms of Service. If you break this policy, you break the Terms.

The rules below give examples. They are not a complete list. If something is harmful, deceptive or illegal, do not do it with RiffAds, even if this page does not name it.

RiffAds does not publish anything for you. It gives you files and download links, and you decide where they go. You are responsible for what you create and for how and where you use it, so review every output before you publish it.

This policy applies to everyone and everything that uses RiffAds through your account or workspace:

• you, and anyone who signs in to your account
• every member of your workspace
• every API key, MCP connection, CLI login and Agent Skill connected to your workspace
• every AI agent, script or tool that acts for you through any of these

If you use RiffAds for a company or for clients, the content you make for them must follow this policy too. Workspace owners and admins must make sure that the members, API keys and connections in their workspace follow this policy.

You are responsible for every action and every charge made with your account, your API keys, your MCP connections and your agents, whether you did it yourself or not.

RiffAds can create custom actors from a photo and clone a voice from a recording. Misused, these features can hurt real people. These rules protect them.

• No deepfakes or impersonation. Do not create content that shows or sounds like a real, identifiable person, or that could make people believe a real person said or did something, unless you have that person's written consent.
• Custom actors. Only use a photo of a real person to create a custom actor (as the actor's photo or as a reference image) if the photo shows you, or a person who has given you written consent to make a digital likeness of them and use it in ads. Having a photo, or finding it online, does not give you this right.
• Voice clones. Only clone your own voice, or the voice of a person who has given you written consent to clone it and use it in ads. Before a voice is cloned, we ask you to confirm that you have this right, and we keep a record of that confirmation.
• Public figures. Do not use the face, voice, name or likeness of a celebrity, influencer, athlete, politician, government official or other public figure without their written consent. This includes lookalikes and soundalikes made to make people think it is them.
• Children. Do not create a custom actor from, or clone the voice of, a real person under 18.
• Respect the limits of consent. Use a person's likeness or voice only in the ways they agreed to. If they withdraw their consent, stop creating new content with it.
• People who have died. These rules also apply to a person who has died, where their family, estate or someone else holds the rights to their name, image or voice.
• No harmful false portrayals. Do not show a real person in a false, humiliating, sexual or defamatory way, including through a lookalike or soundalike.

Keep a copy of every consent you rely on, and show it to us if we ask. If you cannot show it, we may treat the content as a breach of this policy.

An actor or voice from the RiffAds library has not used your product. Do not use one in a way that suggests a real person endorses your product, a cause or a political view.

Ads made with RiffAds must be truthful. They must follow the advertising, endorsement and consumer protection laws of every place where they run, and the rules of every platform where you publish them.

• AI actors are not real customers. Do not present an AI actor as a real customer, reviewer or expert, or a script as a real person's own experience. The rules on testimonials and endorsements differ between countries and platforms, and in some places a label is not enough.
• No fake social proof. Do not invent reviews, ratings, customer counts, sales figures, awards, press mentions or results.
• No fake experts. Do not show an AI actor as a doctor, nurse, pharmacist, scientist, lawyer, financial adviser or other professional in a way that makes viewers believe a real professional recommends the product.
• Health and medical claims. Do not make false or unproven claims that a product treats, cures or prevents a disease or condition, causes weight loss, or replaces medical advice. Only make health claims that you can prove and that the law allows.
• Financial and earnings claims. Do not promise guaranteed returns, easy money or income you cannot prove. Do not advertise financial products, loans, trading or crypto services unless you are allowed to offer them where the ad runs.
• Proof for every claim. You must be able to back up every claim in your ads, including claims you add to a product, a brand kit or a script.
• Clear prices and terms. Do not hide the real price, fees, subscription terms or other important conditions, and do not use bait and switch offers.
• Required disclosures. Include every disclosure the law or the platform requires, for example that an ad is paid, that a person was paid or given something, or that content was made or changed with AI.

See our AI Disclosure for more about AI generated content.

Do not use RiffAds to create, upload, store or share any of the following. This applies to scripts, prompts, uploads, custom actors, voices and outputs.

• Child sexual abuse material, or any content that sexualizes a minor, including an AI actor that looks like a minor.
• Sexual content. Sexually explicit content, including pornography and ads for sexual services, and any intimate or sexual content showing a real person without their consent.
• Hate. Content that attacks, dehumanizes or promotes discrimination against people because of race, ethnicity, national origin, religion, caste, sex, gender identity, sexual orientation, disability, age or serious illness.
• Harassment and bullying. Content that threatens, intimidates, shames or targets a person, or that shares someone's private information without permission.
• Violence and terrorism. Content that threatens, promotes or glorifies violence, or that supports terrorist or violent extremist groups.
• Self-harm. Content that promotes or encourages suicide, self-harm or eating disorders.
• Illegal goods and services, such as illegal drugs, illegal weapons or explosives, counterfeit or stolen goods, fake documents and human trafficking.
• Scams and fraud, such as phishing, fake giveaways, pyramid or Ponzi schemes, fake investment offers, fake charities, and content that pretends to come from a real brand, bank or government body.
• Malware and hacking, including content that spreads malicious software or helps someone break into accounts or systems.
• Anything else illegal where you create it or where it will be shown, or that helps someone break the law.

Child sexual abuse material is never allowed. We will terminate any account linked to it and report it to the relevant authorities as the law requires. To confirm: the process and the authority used to report child sexual abuse material, since no reporting path is set up yet

Some products and services are legal only in some places, only with a license, or only for adults. You may make ads for them only where the product is legal, where you hold every license and approval it needs, and where you follow the advertising rules for it, including age limits and required warnings. This covers, for example:

• gambling, betting, lotteries and sweepstakes
• alcohol
• tobacco, vaping and nicotine products
• cannabis, CBD and hemp products
• prescription medicines, pharmacies, and dietary or weight loss supplements
• weapons, ammunition and weapon accessories
• financial services, loans, trading, investments and crypto assets
• dating services, and adult products that are not sexually explicit

Never aim ads for these products at people under the legal age, and never use an actor who looks like a minor in them. We may refuse or limit content in these categories even where it is legal.

Do not use RiffAds to interfere with elections or political processes. In particular, do not create content that:

• misleads people about when, where or how to vote, who can vote, or election results
• discourages people from voting or tries to suppress votes
• shows a candidate, official, party or government body saying or doing something they did not say or do
• pretends to come from an election authority, a government body or a campaign
• presents AI actors as real voters or supporters to fake public opinion

Any political or issue ad must follow the election and advertising laws where it runs, including rules on showing who paid for it and on labeling AI generated content. To confirm: whether political campaign and issue ads are allowed on RiffAds at all

Only use content you own or have permission to use. Do not upload, import or generate content that breaks someone else's copyright, trademark, publicity or other rights. This includes:

• logos, brand names and trademarks of other companies
• music, sound effects and audio recordings
• product photos, images, videos and ads made by others
• fonts, artwork and characters
• scripts and text copied from other sources

When you import a web page into RiffAds, we read the page and may copy its text and images into your workspace. Only import pages whose content you have the right to use, such as your own website or store.

Do not make ads that copy another brand's look so closely that people may be confused, or that falsely suggest a partnership, sponsorship or endorsement. If you believe content made with RiffAds infringes your rights, write to hello@riffads.com.

You must not misuse RiffAds or the systems behind it. For example, do not:

• copy, decompile or reverse engineer RiffAds, or try to get its source code, except where the law allows it
• scrape or crawl RiffAds, or copy the actor library, the voice library or other content in bulk (for automated access, use only our API, MCP server, CLI and Agent Skills)
• get around or break rate limits, spend limits, plan limits, seat limits or other usage limits
• try to get past our content checks, for example by rewording or splitting a blocked request, hiding text in images or files, or sending a refused request again and again
• access another person's account or workspace, or data you are not allowed to see
• probe, scan or test the security of RiffAds, or run load tests on it, without our written permission
• upload malware, or overload, disrupt or damage the service
• share API keys with anyone outside your workspace, sell, rent or publish them, or share a login instead of inviting members to your workspace
• resell, rent out, sublicense or white label RiffAds, or offer access to it as part of your own product, without a written agreement with us
• create extra accounts or workspaces to get an offer more than once, or to get around limits, a suspension or a ban
• use stolen or unauthorized payment methods, or open chargebacks or payment disputes on valid charges (this may lead to account suspension)
• use RiffAds or its outputs to build or train an AI model or product that competes with RiffAds, or in a way that breaks the terms of the AI model providers whose models power the service

Making ads for your own clients with your workspace is allowed. Selling access to RiffAds itself, or offering it under your own brand, needs a written agreement with us.

RiffAds lets AI agents and your own software create content and spend credits for you, through API keys, MCP connections, the CLI and Agent Skills. An agent that acts through your workspace acts for you.

• Supervise your agents. Give them clear instructions, watch what they do and what they spend, and check their outputs before you publish them. This policy applies to everything an agent creates.
• Their charges are yours. Credits an agent spends are charged the same way as credits you spend yourself, and those charges are final. See our Refund Policy.
• Give only the access needed. If an agent only needs to browse or read, use an API key without spending permissions, or a read-only MCP connection.
• Set your limits. Every agent request that spends credits must state the most credits it may spend. Workspace owners and admins can set a rolling 24 hour credit budget on each API key, and agent spending also has workspace limits per generation and per rolling 24 hours. Limits help, but they do not replace supervision, and you are responsible for setting the limits available to you.
• Keep keys secret. Do not put API keys in public code, in browser code, in shared documents or in prompts that others can see. If a key may have leaked, revoke it at once. A revoked key stops working on its next request, but work that has already started may still finish and be charged. Signing out of the CLI only removes the key from your computer. It does not revoke the key.
• Watch for tricks. An agent can be manipulated by instructions hidden in the web pages, files or messages it reads. You stay responsible for what your agent does, even if someone else tricked it.
• No abuse at scale. Do not use agents to create content that breaks this policy, to test our checks, or to repeat refused requests. We may block a request that is sent again after it was refused.

The AI tools you connect to RiffAds, such as an AI assistant or a code editor, are run by other companies under their own terms. They may keep what they see, including prompts, scripts and links. Check their terms before you connect them.

To help keep RiffAds safe, we use automated checks. They use our own rules and AI models from our service providers. Our Privacy Policy explains how we handle the data involved.

• We check scripts, prompts and other text you enter, image uploads, and some other content, for example the transcript of a voice sample before the voice is cloned.
• For each generation, the checks run before any credits are held for it, so a generation that our checks block uses no credits. In a workflow, each step is checked when it starts, so steps that already ran are still charged.
• The same checks apply in the web app, the API, the MCP server, the CLI and workflows.
• We keep a record of each check and its result.

Automated checks are not perfect. They can miss content that breaks this policy, and they can block content that does not. We do not check every type of content, and we may change our checks at any time. Passing our checks does not mean that content is allowed or legal. You stay responsible for your content.

We may also look into content, accounts and activity when we receive a report, when our checks flag something, or when the law requires it.

If we believe this policy has been broken, or may be broken, we may do any of the following, with or without notice:

• block or refuse a request
• remove content in your workspace, or disable access to it, including uploads, custom actors, voices and outputs
• disable API keys, MCP connections or agent access
• limit features or spending
• suspend your account or workspace, including while we look into a report
• terminate your account or workspace, and refuse new accounts from you
• report illegal content, and the accounts linked to it, to law enforcement or other authorities, and share information with them when the law requires or allows it

We decide what to do based on how serious the problem is, the harm it may cause, and whether it has happened before. We may act at once in cases of serious harm, legal risk or repeated breaches.

We do not give refunds when we act under this policy. All payments are final, credits already used are not given back (including for content we remove), and credits in an account we terminate for breaking our terms are forfeited. See our Refund Policy.

If you think we made a mistake, write to hello@riffads.com and explain why. This section adds to our rights under the Terms of Service.

If you see content made with RiffAds that breaks this policy, or that uses your face, voice, name or work without permission, write to hello@riffads.com. Please include:

• your name and how to contact you
• what the content is and where you saw it, with a link, a file or a clear description
• which rule you think it breaks, and why
• for a likeness or voice complaint: that you are the person shown or heard, or that you are allowed to act for them
• for a copyright or trademark complaint: the work or mark you own, and proof that you own it or act for the owner

We may ask for more information to confirm who you are or what rights you hold. We review reports and act as this policy allows. Do not send reports you know are false.

RiffAds does not publish content anywhere. If you found the content on a website or a social platform, report it to that platform too, because only they can remove it there.

If you find child sexual abuse material, do not download it, copy it or send it to us. Report it to the police or to a child protection hotline in your country, and tell us where you found it.

We may update this policy from time to time. The date at the top of this page shows when it last changed. If you keep using RiffAds after a change, the updated policy applies to you.

Questions about this policy? Write to hello@riffads.com.