You are using AI: the AI Booking features labelled “AI assistant”, “chatbot” and similar generate answers or text with artificial intelligence. Do not treat the result as verified fact — always review it with human judgement before important decisions.
1. Affected features
- AI chatbots: the providers’ booking chatbot and the aibooking.hu information chatbot give conversational answers about booking and services.
- AI assistant: a helper available in the provider and guest mobile apps that gives informational answers based on the question and the minimum context needed to operate.
- AI helper features in the provider interface: service import (from a given link or price-list text), price suggestions, no-show risk estimation, drafting replies to reviews and description generation.
AI answers are produced by the language models of Ollama Cloud and OpenAI; both are data processors. The interface clearly indicates the AI nature at the start of the interaction. Details of the AI providers and data handling are set out in the privacy policy.
2. Limitations and human oversight
AI can be wrong, invent information, misunderstand the question or produce unrealistic details. The generated result does not constitute legal, financial, medical, safety or professional advice. Human review is required before changing settings, making contractual decisions, taking steps affecting guests’ rights or publishing publicly.
3. Labelling AI content
Where the AI talks directly with the guest (chatbot, assistant), the interface indicates at the start of the conversation that artificial intelligence is answering. Where a provider publishes text drafted with AI (for example a review reply or a description), the provider is responsible for human review of the content and — where needed to avoid deception — for keeping a label such as “created / assisted by artificial intelligence”.
4. Prohibited and unsupported use
- biometric identification, categorisation of a person or inferring their sensitive characteristics;
- making employment, credit, insurance, healthcare, law-enforcement or other significant decisions based on AI output without human review;
- deceiving, harassing, impersonating, defaming or unlawfully manipulating another person;
- use to the detriment of a minor or a vulnerable person;
- creating unlawful, sexual, hateful, violent content or content that infringes others’ rights.
5. Data and confidential information
Do not enter passwords, PIN codes, API keys, bank card data, special-category (e.g. health) data or unnecessary guest data into the AI features. The system only forwards the context needed for the answer to the processor. According to Ollama and OpenAI, input and output submitted via the API are not used for model training by default.
Before using the AI features we ask for separate, explicit consent; data is forwarded to the processor only after acceptance. Further retention is governed by each provider’s own privacy terms.
6. EU AI Act compliance
Taking into account the transparency requirements of Article 50 of Regulation (EU) 2024/1689 (AI Act), applicable from 2 August 2026, AI Booking signals AI interaction and artificial output. We update the technical labelling options as the relevant standards, codes of practice and provider capabilities evolve.
7. Who is responsible for what
AI Booking, the developer and operator of the AI features, provides the AI signal at the start of the interaction, names the model providers, asks for consent before the AI features are used, and enables human oversight.
The provider who turns on the AI chatbot or assistant towards their own guests, as the party deploying the AI system, is responsible for informing their guests, having the generated content reviewed by a human before publication, and keeping the labelling in their own interface and communication.
8. Feedback and rights
Incorrect, harmful or misleading AI output can be reported at info@aibooking.hu. For data protection rights and deletion requests, the privacy policy applies. AI Booking may restrict the AI feature for security or compliance reasons.