Legal

AI Transparency Notice

Last updated: 25 June 2026  ·  Magneety EU Ltd.

Magneety uses artificial intelligence to help you create and improve your marketing. This notice explains, in plain English, where we use AI, which AI we use, and how we meet our transparency obligations under the EU AI Act (Regulation (EU) 2024/1689), in particular Article 50. It sits alongside our Privacy Policy and Terms of Service.

1. How Magneety is classified under the EU AI Act

• Not a prohibited practice. Magneety does none of the AI uses banned under Article 5 (see Section 7).
• Not high-risk. Magneety generates marketing content and recommendations. It is not used in any of the high-risk areas listed in Annex III of the Act (such as employment, credit scoring, biometrics, law enforcement, or critical infrastructure). It is therefore not a high-risk AI system.
• Limited-risk - transparency applies. Because Magneety includes an AI assistant you chat with and generates synthetic content (text, images, video), it falls under the Article 50 transparency obligations. This notice, plus the in-app AI labels described below, are how we meet them.
• Our role. Magneety is a deployer of third-party general-purpose AI models and a provider of the AI features built on top of them. The underlying model providers carry the obligations that apply to the models themselves.

2. Where Magneety uses AI

• Content generation - captions, headlines, ad copy, email and landing-page text, quizzes.
• Image and video generation - AI-created visuals for posts and ads.
• Summaries and insights - plain-English explanations of your performance data.
• Reply suggestions - draft replies to comments, messages, and reviews.
• Recommendations - suggested campaign settings, budgets, and next actions.
• The Magneety assistant - the in-app chat that answers questions about your marketing.
• AI Search Visibility (GEO/AEO) - checking whether AI assistants mention your brand, and drafting content to improve that.
• Autopilot (optional, off by default) - automated budget and pacing suggestions you can choose to let it apply, within limits you set.

3. Which AI we use

We use established third-party AI providers. We do not train our own foundation models, and your data is not used to train theirs:

• Anthropic (Claude) - text generation, summaries, reply suggestions, and the assistant. Covered by Anthropic's Commercial Terms; your inputs are not used to train Anthropic's models.
• Replicate and FAL.ai - AI image and video generation.
• OpenAI - limited use for location lookups in ad targeting.

These are general-purpose AI providers. They are listed as sub-processors in our Privacy Policy, with international transfers covered by Standard Contractual Clauses.

4. How we meet our transparency obligations (Article 50)

• You always know when you are talking to AI. The Magneety assistant is clearly identified as AI, every time you use it. It never pretends to be a human.
• AI-generated content is labeled. Content produced by AI inside Magneety carries a clear "AI-generated" label so you can tell it apart from anything you wrote yourself.
• Generated images are marked. We are rolling out machine-readable marking of AI-generated images (the format that lets other tools detect them as AI-created) ahead of the Article 50(2) deadline of 2 December 2026. Until then, AI images are visibly labeled in-product.
• Plain English. We explain what the AI is doing in language a non-expert can follow, so you can make informed choices (the Act's AI-literacy aim).

5. You stay in control - human oversight

Autopilot is off by default. If you turn it on, it acts only within the limits you set (such as budget caps), every action is logged, and you can undo it. You can pause or disable it at any time.

6. Accuracy and limitations

AI can be wrong, out of date, or biased, and can produce content that sounds confident but is inaccurate. Always review AI output before you publish or act on it. AI suggestions are not professional, legal, financial, or medical advice. You are responsible for the content you choose to publish.

7. What Magneety does not do

We do not use AI for any practice prohibited by the EU AI Act or that we consider harmful. Magneety does not:

• Score or rank people socially.
• Use subliminal, manipulative, or deceptive techniques, or exploit vulnerabilities, to distort behaviour.
• Perform biometric categorisation or emotion recognition.
• Build facial-recognition databases or scrape faces from the internet.
• Make consequential decisions about people automatically, without a human in the loop.

8. Your responsibilities when you publish

When you publish AI-generated content to your own audience, you may have your own transparency duties under the AI Act, advertising rules, and consumer-protection law - for example, disclosing AI-generated or manipulated images of real people. Magneety gives you the tools (labels, machine-readable marking, human review), but the choice of what to publish, and meeting any disclosure duty that applies to you as the publisher, is yours. We recommend always keeping a human in the loop and disclosing AI content where your audience would reasonably expect to know.

9. Data and AI

To generate content we send the AI providers above only what is needed for the task (for example your prompt, brand voice, and relevant product or performance context). This is governed by our Privacy Policy and, where personal data is involved, by the GDPR. Your data is not used to train the providers' models.

10. Key dates

• August 2025 - rules for general-purpose AI models began to apply (these sit with the model providers).
• 2 August 2026 - the Article 50 transparency obligations become enforceable.
• 2 December 2026 - deadline for machine-readable marking of AI-generated content already in the market.

11. Questions, concerns, or human review

If you have a question about how we use AI, want a human to review an AI-related outcome, or want to report a concern, contact us at info@magneety.com. We respond within 24 hours on business days.

12. Changes to this notice

As the EU AI Act guidance and codes of practice are finalised, and as our AI features evolve, we will update this notice. The "Last updated" date above reflects the most recent revision. This notice is provided for transparency and is not legal advice.

Magneety EU Ltd.  ·  5 Ekzarh Yosif I Str., 2nd floor, Office 1, 9300 Dobrich, Bulgaria