1. Why this page exists
VinoSomm deliberately uses artificial intelligence (AI) - in the design of this website and in core features of the app. This page discloses that use openly and fulfils the transparency obligations of the EU AI Act (Regulation (EU) 2024/1689, in particular Article 50).
2. AI-generated content on this website
The photographic and video imagery on this website (in particular the black-and-white scenes and portraits) is generated with AI image and video models or edited with AI assistance. The people depicted do not exist; any resemblance to real persons would be coincidental and is not intended.
Parts of the texts on this website are also created with AI assistance and are reviewed editorially before publication. The app views shown are demonstrations with sample data; they do not show data of real users.
3. AI features in the VinoSomm app
Sommelier chat (AI assistant)
In the sommelier chat you are talking to an AI system, not a human being. The chat is labelled accordingly in the app. Despite careful development, AI answers can be incomplete or incorrect; they are recommendations for orientation, not binding advice.
Your taste profile
Your taste profile is computed by AI models from your ratings, tastings and your use of the app, and is refined continuously. It is an algorithmic estimate of your preferences, not an objective measurement.
Wine profiles in the catalogue
The sensory profiles of the wines (aromas, structure, descriptions) are created and curated with AI assistance from publicly available information, in particular from the wineries. Despite quality assurance, individual details can be incomplete or incorrect; the producer's information on the label always prevails.
Match scores and recommendations
Match scores result from algorithmically comparing your taste profile with the wine profiles. They are an estimate of how well a wine is likely to suit you, not a guarantee.
Scan features
When you scan labels, wine lists or dishes, AI image recognition is used to identify wines and to suggest suitable recommendations.
4. Providers we use
For AI features we work with specialised providers (currently OpenAI, Mistral AI and Anthropic). Details on data processing, third-country transfers and the contractual commitment that your content is not used to train their models can be found in our privacy policy.
5. Limits and responsibility
AI systems can make mistakes. We continuously improve our models and review processes, but we cannot guarantee the accuracy of individual AI outputs. If you notice an incorrect or inappropriate output, please let us know at info@vinosomm.ai.
Last updated: 13 August 2026. We will update this page whenever our use of AI changes materially.