Avocado AI Content Policy

Last updated: September 28, 2026

This policy explains what you may create, publish, and automate with Avocado AI. It applies to everything you do in the app, in Super Agent, in Flows, and through the MCP server. It works together with our Terms of Service. If the two differ, the Terms apply.

How filtering works

Your prompts, uploads, and outputs are checked by our systems and by the AI provider that runs the model you picked. If something is blocked you will see a message such as "content policy violation" or "flagged material". That means the request broke a safety rule, not that the system failed. Credits for a blocked generation are returned.

Never allowed

• Sexual content involving minors, or anyone who looks under 18, in any context.
• Pornography, explicit nudity, "undressing" or clothing-removal edits, and non-consensual sexual content.
• Realistic gore, torture, or content that glorifies violence.
• Hate speech, extremist propaganda, or harassment of people or groups.
• Content that promotes self-harm, suicide, or eating disorders.
• Help with weapons, drugs, fraud, or other illegal activity.
• Election manipulation, fake news footage, or content presented as a real recording of events that did not happen.
• Malware, phishing pages, scam scripts, or impersonation of companies or public bodies.

Real people, faces, and voices

You may create content with real people only when you have the right to. That means:

• Yourself: you can use your own photos and voice.
• People who agreed: you can use photos, video, or voice recordings of people (for example your creators, staff, or models) only with their explicit consent to AI generation.
• Consent confirmation: some features, such as AI UGC with real faces, ask you to confirm consent before each use. We record that confirmation. Confirming consent you do not have breaks our Terms.
• Never: celebrities, politicians, influencers, or private individuals without consent; voice clones of anyone who did not agree; deepfakes meant to deceive.

Brands, products, and intellectual property

• Your own brand: your logo, products, packaging, and brand assets are welcome. That is what Avocado is built for.
• Other people's brands and characters: do not generate other companies' logos, products, characters (for example Disney or Marvel characters), or packaging, and do not imply endorsement.
• Music and voices: do not imitate a named artist's voice or copy lyrics or melodies you do not own. Describe mood and style instead.

Advertising and claims

• Do not make false or unprovable claims about health, medicine, money, or results.
• Follow the advertising rules of the platform you publish to (for example Meta's advertising policies) and your local law, including labeling of ads and AI-generated content where required.
• Do not use generated "customer testimonials" or before-and-after images that suggest real results unless they are real.

Publishing, automations, and messages

When Super Agent publishes, runs a comment-to-DM automation, or launches ads for you, you are the publisher. You must:

• Only message people who interacted with your post or otherwise expect to hear from you. No spam or bulk unsolicited messages.
• Keep automated messages honest and not misleading.
• Follow each platform's rules. Platforms can restrict accounts that break them, and we cannot reverse that.

Transparency

Do not present generated media as real footage of real events. Where a platform or law requires an AI label, add it.

Enforcement

Blocked requests are not charged. Repeated or deliberate violations can lead to suspension of generation, removal of content, or account termination. We may report illegal content to the authorities.

Appeals and reports

If you think a request was blocked by mistake, email support@avocadoai.co with the subject "Content Policy Appeal", the prompt, the time, and a short explanation. To report content made with Avocado that harms you, email legal@avocadoai.co.

Quick checklist

• Does my content show a real person? Do I have their consent?
• Does it use someone else's brand, character, or music?
• Could it be mistaken for a real event or real result?
• Would this message or ad be welcome and honest to the person who receives it?