Version: July 2026
Medioora offers optional AI functions for creating, revising and reviewing requirements, architecture and design descriptions, and test cases. The AI may also provide review and improvement suggestions.
Explicit activation
An AI request is sent only after the user selects the AI function and submits a prompt. No communication with the AI provider occurs without this action. Administrators can disable the function for their organisation.
Data sent
Only the prompt, the document currently open and item-level context expressly selected by the user are sent. Other project content is not transmitted automatically. Users must not submit unnecessary personal data, special-category data, patient data or confidential information without authorisation and suitable safeguards.
Provider and safeguards
Medioora uses the OpenAI API. OpenAI processes requests as a technical processor under the applicable data processing agreement (DPA/AVV). API settings are configured so that inputs and outputs are not authorised for training or improvement of OpenAI’s general models, subject to the applicable contract and configuration. Security, abuse-monitoring or legally required logs may still occur. The current model is from the GPT-5 family.
Human review
AI output can be incomplete or wrong. It is a suggestion, not legal, medical, regulatory, safety or other professional advice. Users must review it before relying on it. AI output must not be the sole basis for decisions producing legal or similarly significant effects.
Changes
AI models, providers and availability may change or be discontinued for security, quality or contractual reasons. Questions can be sent to info@medtech-ingenieur.de.