Terms of use and copyright

1. Who provides the site and tools

The website hypervision.si and the Clean before AI browser extension are created and run by David Radovič from Slovenia as an independent project. Contact: hello@hypervision.si.

These terms apply to the whole site, its tools and the extension. By using them you agree to these terms.

2. Your content stays yours

• Texts and files you process with the tools remain yours. We acquire no rights in them — processing happens in your browser and we never see them (see the Trust Center).
• You may freely use the results: cleaned texts, redacted documents and documents created with the business tools (for example an AI use notice or an AI system register).

3. What you may do

• Use the web tools and the extension free of charge for personal and business purposes, in the form in which they are published.
• Use prompts from the library in your work, adapt them and share individual prompts with colleagues.
• Learn in the AI Academy — on your own or with your team, through this site.
• Share links to the site and quote short excerpts with attribution (hypervision.si).

4. Not allowed without written permission

• copying, modifying, disassembling or imitating the code, detection engine (rules, patterns, checksums), model and dictionaries of Clean before AI, or building them into other products or services;
• extracting the model, dictionaries or rules from the extension or the site and using them separately;
• using the product, its outputs or parts of it to train or improve other products for detecting or anonymising personal data;
• systematically downloading the site’s content or substantial parts of its collections (prompt library, tool radar, overview of regulations, lessons), or republishing, selling or using them to train AI models;
• distributing, selling or renting the extension or parts of it — the extension is installed only from the Chrome Web Store;
• removing, altering or obscuring copyright and trademark notices or other information about origin and rights management;
• loading the site or tools in a way that disrupts their operation.

These restrictions do not affect rights the law expressly grants you (for example quotation or the rights of a lawful user of a computer program), nor the licences of the open-source components listed in section 8.

5. Copyright, collections and trademarks

• © 2026 David Radovič. All rights reserved. This covers the code, detection engine, model, dictionaries, texts, prompts, lessons and quizzes, graphics, videos and the design of the site.
• The collections on the site — prompt library, tool radar, overview of regulations and name dictionaries — are also protected as databases.
• Clean before AI™ and Hypervision™ are trademarks of David Radovič.
• We reserve the right of text and data mining (Article 4(3) of Directive (EU) 2019/790): the content of the site and the products may not be used to train AI models without permission.
• Third-party content (for example news headlines and excerpts, or the names and logos of tools) belongs to its owners.

6. No warranty

• Clean before AI reduces risk; it does not guarantee complete protection. Automatic detection may miss an item or mark too much, so review the result before sending.
• Content about regulations, tool ratings and recommendations is general information, not legal or other professional advice.
• The tools and content are free and provided “as is”, without any warranty that they will always be available, free of errors or fit for your purpose.

7. Liability

To the fullest extent permitted by law, we are not liable for damage arising from the use of, or inability to use, the site, the tools and the extension, including damage caused by data that was not detected. This does not apply to damage caused intentionally or through gross negligence, or where the law does not allow liability to be limited.

8. Third parties and open-source components

• AI services (ChatGPT, Claude, Gemini, Copilot …) are provided by third parties; their own terms and privacy policies apply.
• Links to other websites are for information; we are not responsible for their content.
• The site uses open-source libraries, which remain under their own licences: Next.js, React (MIT), pdf.js (Apache-2.0), pdf-lib (MIT), Mammoth (BSD-2-Clause), JSZip (MIT), Tesseract.js (Apache-2.0), Transformers.js (Apache-2.0), ONNX Runtime Web (MIT), heic-to (LGPL-3.0), exifr (MIT), jsQR (Apache-2.0), node-qrcode (MIT), jsdiff (BSD-3-Clause).
• Models that some tools download from Hugging Face on first use (for example GLiNER, Whisper, Qwen, Gemma, Granite) belong to their authors and are subject to their licences.

9. Changes

We may update these terms, for example when new features are added. The new version applies from its publication on this page; the date is shown above.

10. Governing law and jurisdiction

These terms are governed by the law of the Republic of Slovenia. Disputes fall under the jurisdiction of the competent court in Novo mesto.

If you are a consumer, this does not deprive you of the protection of the mandatory laws of the country where you live, or of your right to bring proceedings before the courts where you live. If any provision is invalid, the remaining provisions remain in force.

11. Contact and permissions

Questions and requests for permission — for example to build the Clean before AI engine into your systems or to reuse content: hello@hypervision.si.