Avenlo Avenlo
Home
Dashboard Profile
Logout
Sign In Sign Up

AI Disclosure

Last Updated: 6 November 2026 · How Avenlo uses artificial intelligence and what that means for you and your audience

1. Avenlo generates content with AI

Most of what Avenlo produces is made with artificial intelligence. Depending on the features you use, AI systems write topics, scripts, captions, and post copy; synthesise narration (including cloned voices); generate images and background visuals; transcribe audio for subtitles; and assemble these into finished videos. We use third-party AI providers for this — currently Anthropic (text), ElevenLabs (speech and voice cloning), and OpenAI / Replicate (images). They are listed with what they receive in our Privacy Policy.

2. AI output can be wrong

AI-generated text, speech, and images can be inaccurate, outdated, biased, or unintentionally similar to existing works. You are responsible for reviewing every piece of content before it is published and for any claims it makes. Do not rely on AI output for medical, legal, financial, or safety-critical information without checking it.

3. Labelling AI-generated content

The EU Artificial Intelligence Act (Regulation (EU) 2024/1689, Article 50 — applicable since 2 August 2026) and the rules of TikTok, YouTube, and Instagram require that certain AI-generated or AI-altered content is disclosed to the people who see it:

  • Realistic synthetic media ("deepfakes") — video, audio, or images that look like a real person, place, or event — must be labelled as AI-generated or manipulated.
  • Platform labels. When Avenlo publishes a video made with its AI pipeline (Studio or Auto‑Pilot), it turns on the platform's own AI label: TikTok's "AI-generated content" label and YouTube's "altered or synthetic content" disclosure. Instagram doesn't let apps set its AI label, so on Instagram the label is applied by Instagram itself or by you in the Instagram app. Videos you upload yourself are not labelled automatically; if they contain realistic synthetic media, you must label them on the platform.
  • Machine-readable marking. We are adding provenance metadata (C2PA-style content credentials) to videos rendered by Avenlo so that they can be recognised as AI-generated by software. Do not strip this metadata.

Turning a required label off, or publishing undisclosed deepfakes, violates our Acceptable Use Policy and may violate platform rules and the law.

4. Automation: Auto‑Pilot and automatic replies

Auto‑Pilot is an opt‑in feature that autonomously generates AND publishes content to the social accounts you select, on the schedule and with the settings you configure — without asking you to approve each post. Nothing is published this way unless you have turned Auto‑Pilot on; you can pause or disable it, and change its cadence, topics, and target accounts, at any time. Content published by Auto‑Pilot is AI-generated, is published in your name and on your responsibility, and carries the AI labels described in section 3.

If you enable automatic replies to comments or direct messages, the people who receive them are interacting with an automated system, not with you in real time. Avenlo therefore marks automated replies as such (for example with an "automated reply" note, and in direct messages at the start of the conversation), as required by the EU AI Act, Meta's messaging policies, and consumer-protection law. You must leave this disclosure in place. A human can always take over the conversation from your inbox.

5. Voice cloning

Voice cloning creates a synthetic voice from audio samples. You may only clone your own voice or a voice whose owner has given you explicit, documented consent. Your voice sample and the resulting voice model are processed by ElevenLabs on our behalf; we treat the sample as biometric data and process it only with your explicit consent, which you can withdraw by deleting the voice clone in your settings. Content produced with a cloned voice that resembles a real person must be labelled as synthetic (section 3).

6. What we do not do

  • We do not use your content, your voice, or the comments and messages of your audience to train AI models, and our AI providers are contractually bound not to train on the data we send them.
  • The Avenlo Assistant (the chat button in the app) is an AI, not a person. We keep your conversations with it for up to 90 days to answer support requests and improve Avenlo, never to train AI models; see the Privacy Policy.
  • We do not send the names, e-mail addresses, or social-media handles of you or your audience to AI providers; prompts contain only the text, scripts, transcripts, and video frames needed to generate the requested output.
  • We do not make automated decisions about you that have legal or similarly significant effects. Account restrictions for policy violations are reviewed by a person.

7. Questions

Contact [email protected]. For questions about personal data, see the Privacy Policy.

Legal Pages

Terms of Service · Privacy Policy · Impressum · Refunds & Cancellation · Right of Withdrawal · Cancel Contract · Acceptable Use · AI Disclosure · Report Content · Data Processing Agreement · Data Deletion