Skip to content
Back to Feeverso

Placeholder legal entity — replace before launch

Values shown in [brackets] come from src/config/legal.ts and must be filled in before this page goes live.

Legal

AI Content Policy

Last updated: October 24, 2026

Feeverso is an AI marketing workspace. The agent chat, the Studio templates and the media generator all produce content with third-party artificial intelligence models. This policy explains how those models are used, what you can and cannot expect from their output, what you must not create, and what you must disclose when you publish. It applies to every plan and forms part of our Terms of Service.

In short: the AI drafts, you decide. You are responsible for what you publish.

1. Which AI models Feeverso uses

Text. Feeverso accesses text models through OpenRouter, Inc. OpenRouter routes each request to a model provider. Depending on your plan and the task, requests may be handled by providers such as DeepSeek, Anthropic, OpenAI or Google. The models available on your plan are listed in the app and on our pricing page.

Images and video. Feeverso accesses image and video models through Atlas Cloud. These models generate still images, product-photo edits and short videos from your prompts and, where you upload them, your own reference images.

We do not build or train foundation models. We select models, route requests, apply templates and prompts, and present the results to you. Model providers process your prompts under their own terms as well as under our agreements with them. Our Privacy Policy describes the data categories involved and the safeguards that apply to transfers outside the EEA.

2. What the AI can and cannot do

AI models are probabilistic. They predict plausible output; they do not check facts and they do not know your business.

  • Output may be inaccurate or incomplete. Models can invent statistics, prices, product features, quotations, dates, sources or legal references. They can also state outdated information with confidence.
  • Output may be similar to content generated for others. The same prompt can produce similar results for different users. We do not guarantee that output is unique, original or free from third-party rights.
  • Output may be unsuitable for your market. Tone, humour, idioms, cultural references and regulatory requirements differ between countries.
  • Visual output may contain errors. Image and video models can distort text inside an image, hands, faces, reflections, product labels, packaging and logos. Short videos may show inconsistent characters or objects between frames.
  • We do not guarantee performance. We make no promise about conversion rates, advertising approval, search ranking, reach, engagement, revenue or any other commercial result.
  • Models change. Providers update, deprecate or restrict models. A prompt that worked last month may produce a different result today, and a model may be withdrawn from your plan.

3. You must review everything before you publish

Human review is required. Before you publish, send or use any output, check at least:

  • facts, figures, prices, dates, names, product specifications and availability;
  • claims about health, safety, performance, income, savings or environmental impact;
  • statements about competitors and comparisons;
  • legal, medical, financial and tax statements;
  • third-party trademarks, logos, packaging, characters and locations;
  • the likeness of any person shown or imitated;
  • local advertising, consumer-protection and platform rules.

You are the publisher. As between you and Feeverso, you are responsible for the content you publish, for the permissions you need, and for complying with the advertising, consumer-protection, intellectual-property and competition laws that apply to you.

4. Prohibited uses

You must not use Feeverso to create, edit or publish content that:

  • depicts a real person without their consent, including a likeness, voice or mannerism used to suggest that the person said or did something they did not do;
  • is a deepfake intended to deceive, or that presents a real person or real event in a false or misleading way;
  • impersonates a person, brand, public authority or organisation, or falsely suggests an endorsement, partnership or affiliation;
  • contains fake reviews, fake ratings or fake testimonials, or presents invented customers as real;
  • makes misleading health claims, including claims that a product diagnoses, treats, cures or prevents a disease without adequate substantiation;
  • makes misleading financial claims, including guaranteed returns, risk-free investments, get-rich-quick schemes or unlicensed financial advice;
  • misuses trademarks, copyright, design rights or trade secrets, or passes off someone else's work as your own;
  • is sexual or sexually suggestive, or sexualises any person;
  • involves or depicts minors in a sexual, harmful or exploitative way;
  • harasses, threatens, defames or promotes hatred against a person or group;
  • spreads disinformation about elections, public health, emergencies or other matters of public interest;
  • promotes illegal goods or services, or facilitates fraud, scams or deception;
  • violates privacy, for example by publishing someone's personal data without a lawful basis;
  • attempts to bypass safety filters, or uses the service to build or train a competing AI model.

We may suspend or close an account that repeatedly or seriously breaches this section, and we may report unlawful content to the competent authorities where the law requires it.

5. Real people, likenesses and deepfakes

You may use a real person's image, voice or likeness only if you have that person's informed consent and the use is lawful. Consent must cover the specific use, including paid advertising and the channels where you will publish.

Never use Feeverso to create:

  • sexual or intimate content involving a real person, whether or not they are famous;
  • content that puts words into a real person's mouth or shows them doing something they did not do;
  • content involving a real minor;
  • political or election content that falsely presents a candidate or official;
  • content that could reasonably be mistaken for authentic footage of a real event.

If a person withdraws consent, stop using the content and remove it from the channels you control.

6. Reviews, testimonials and endorsements

  • Only real customers, with permission. Use a testimonial only if the person or business actually said it and agreed to its use.
  • Do not invent reviews, ratings, case studies or results. Inventing a customer or a result is a deceptive commercial practice in the EU and a deceptive endorsement in the US.
  • Disclose incentives. If you gave a free product, a discount or payment for a review or post, say so clearly and close to the content.
  • Keep the meaning intact. Do not edit a testimonial so that it says something the person did not mean.

7. Health, financial and other regulated claims

Marketing content often touches regulated areas. Extra care is required for:

  • health and supplements — no claims to cure, treat or prevent disease without the substantiation and authorisations the law requires;
  • finance and crypto — no guaranteed returns, no misleading risk statements, no unlicensed investment, credit or insurance advice;
  • alcohol, gambling and tobacco — comply with the age limits, warnings and targeting restrictions that apply in each market;
  • children and education — no manipulative or age-inappropriate messaging;
  • employment and housing — no discriminatory wording in job adverts or property listings.

AI output is a draft, not professional advice. Feeverso does not provide legal, medical, tax or financial advice, and nothing generated by the service should be relied on as such.

8. Disclosure and labelling

You must tell people when content is AI-generated where the law or the platform requires it.

  • EU AI Act, Article 50. Deepfakes — images, audio or video of real people or real events that are artificially generated or manipulated — must be disclosed as such, clearly and in a way that is noticeable at the latest at the point of first exposure. AI-generated text published to inform the public on matters of public interest must also be disclosed, unless a human has taken editorial responsibility and the law provides an exception.
  • Platform rules. Meta, TikTok, YouTube and other platforms require creators to label realistic AI-generated or altered content. Use the platform's own AI-content label or toggle where it applies, and do not remove labels that a platform has applied.
  • Sponsored content. Mark paid partnerships and advertising as such, using the platform's paid-partnership tool and a clear label such as "#ad" where that is the local convention.
  • Do not misrepresent the source. Do not present AI-generated content as authentic photography, real customer footage or an unedited recording when that would mislead.

A label does not cure a misleading message. Disclosure is required in addition to — not instead of — telling the truth.

9. Safety filters and blocked requests

Model providers and Feeverso apply safety filters. A request may be refused, or output may be withheld or cut short, for example because it appears to involve a real person, sexual content, violence, self-harm, hate, illegal activity or a minor.

  • Filters can be wrong in both directions. A legitimate marketing request can be blocked, and a problematic one can occasionally get through.
  • Failed generations are refunded. If a generation fails, the credits are returned to your balance automatically. Credits are not lost because a model refused a request.
  • Do not try to evade the filters. Repeated attempts to work around a refusal, for example by rephrasing to hide the real subject, are a breach of this policy.

10. Ownership and licences

  • Your inputs stay yours. You keep all rights in the text, images, brand kits and other material you upload.
  • You own your outputs. As between you and Feeverso, and to the extent the law permits, you own the content you generate. We assign to you any rights we may have in that output.
  • We get a limited licence. You grant us a non-exclusive, worldwide licence to host, store, reproduce, process, transmit to model providers and display your content only as needed to run the service for you — including generating the output, showing it in your account, creating previews and backups, and complying with the law. This licence ends when you delete the content or your account, apart from backups and records we must keep.
  • You may use your output commercially, subject to this policy, our Terms of Service and the rights of third parties. You are responsible for clearing any third-party material that appears in your output.
  • No exclusivity. Because models are shared, we cannot guarantee that similar output is not generated for another customer, and we do not grant exclusivity in any prompt, style or result.

11. No training on your content

Feeverso does not train AI models on customer content. We do not use your brand kits, prompts, uploads, conversations or generated assets to train our own models or third-party models. Your content is sent to a model provider only to fulfil the request you made.

You must not use Feeverso output to train, fine-tune or distil a competing model, or to build a dataset for that purpose.

12. Reporting a problem

If you see content on Feeverso that you believe is illegal, deceptive or breaches this policy, or if you believe someone has used your likeness or work without permission, email support@feeverso.example with:

  • the account or generation concerned, or a link or screenshot;
  • what the problem is and why;
  • your contact details and, if you are the rights holder, evidence of your rights.

We review reports, may remove content and may suspend accounts. We aim to reply within five business days. If you are in the EU, you can also report illegal content to your national Digital Services Act authority.

13. Changes to this policy

We may update this policy as the models, the law or the platforms change. We announce material changes in the app before they take effect. The version in force when you generated content governs that content. Nothing in this policy limits rights you have under mandatory consumer law.