AI Content and Liability
Rights in generated creatives, their accuracy, and your review obligations before publishing.
Effective date: 2026-09-09
This document defines the checks you must carry out yourself before publishing content generated with Adropic. If you publish without performing them, all resulting liability is yours.
This document forms an integral part of the Terms of Service and is read together with sections 15, 16 and 17 of that agreement.
1. Rights in Output
Adropic assigns to you all rights it holds in generated Output, to the extent applicable and provided your payments are current.
However, the scope of copyright protection for AI-generated content differs by country. In many jurisdictions, output without human authorship is not eligible for copyright protection at all. Adropic gives no warranty as to the registrability, protectability or exclusivity of Output.
2. No uniqueness guarantee
Identical or similar inputs may produce similar Output for different users. Adropic does not undertake that any Output has not been generated for another user, and has no obligation to prevent third parties from using similar Output.
3. Third party rights and infringement risk
AI models may, as a result of their training data, produce Output resembling the copyright, trade mark or design rights of third parties.
Adropic does not warrant that Output is free of third party rights and has no obligation to defend or indemnify you in any resulting claim. On the contrary, if such a claim arises you are obliged to indemnify Adropic under section 17 of the Terms of Service.
Before publishing, you must review Output for similarity to trade marks, logos and well-known designs.
4. Accuracy and hallucination risk
Generated images, videos and text may deviate from your product's real appearance, material, colour, packaging, size or features. The model may generate a feature, label, certification or packaging text that does not exist.
Reviewing every creative before publishing is your obligation. A product image that does not reflect reality is treated as a misleading commercial practice in many jurisdictions and can result in administrative fines.
What to check:
- Does the product's colour, form, packaging and label match reality
- Does the image show an accessory, gift or second product that is not included
- Are the price, discount rate and campaign duration in the copy correct
- Has the image rendered a fabricated certificate, award or badge as legible text
5. Claims in copy
You must verify the accuracy of performance, health, discount and comparative claims in generated ad copy. Phrases such as "best", "100%", "clinically proven" or "guaranteed results" must not be used without documented substantiation.
Copy generated in the following areas must be separately assessed against applicable law before publication:
| Area | Typical risk |
|---|---|
| Health, supplements, cosmetics | Therapeutic claims, unapproved indications |
| Finance, crypto, investment | Return promises, missing risk warnings |
| Weight loss, aesthetics | Misleading before/after imagery |
| Comparative advertising | Unfair reference to a competitor's brand |
| Products aimed at children | Age-inappropriate presentation |
6. AI disclosure and labelling
Some platforms and jurisdictions require AI-generated advertising content to be labelled. You are responsible for applying that label; Adropic does not add a disclosure label to Output automatically.
7. Likeness, voice and avatars
Generated avatars and characters are fictional. Output may nevertheless resemble a real person. Where a real person's face, voice or likeness is involved, you must have obtained that person's documented, explicit consent; see the Acceptable Use Policy.
8. Competitor analysis output
Content derived from competitor brand analysis may be used for analysis and inspiration only. Directly copying and publishing a competitor's ad imagery, slogans or copy may constitute unfair competition and copyright infringement; liability for such use is yours.
9. Third party models
Generation takes place through third party model providers. Those providers' own usage policies apply to the content you generate. If you breach a provider's policy, generation may be blocked or your account suspended.
Adropic may change its model providers without prior notice. A change of model may affect the style and quality of Output; this does not constitute a defect in the service.
10. Automated moderation
Adropic applies automated checks that block clearly unlawful or policy-violating generation requests. These checks are not exhaustive; the absence of a block does not mean the generated content complies with the law or with platform policies.
11. Summary of responsibility
Adropic is a generation tool, not a publisher or advertising agency. You decide what gets published. Administrative, civil and criminal liability for published content is yours.