M.O.O.N. – My Opposite of Now

This Privacy Notice explains how personal data is processed in connection with the use of an interactive, web-based AI-powered application provided by the Jewish Museum Berlin Foundation. The application can be accessed at a media station in the exhibition The Opposite of Now and via the website of the Jewish Museum Berlin: www.jmberlin.de/moon.

Last updated: August 2026

Contents

1. Controller
2. Data Protection Officer
3. General Information on Data Processing 
4. Categories of Data Processed 
5. Purpose of Processing 
6. Legal Bases for Processing 
7. Processing of Chat Content 
8. Publication of Content 
9. Use of the AI Model 
10. Hosting 
11. Recipients of Personal Data 
12. Transfers to Third Countries 
13. Data Retention
14. Access Rights 
15. Data Subject Rights 
16. Right to Lodge a Complaint 
17. Automated Decision-Making 
18. Data Security 
19. Changes to this Privacy Notice

Controller

The controller within the meaning of the General Data Protection Regulation (GDPR) is:

Stiftung Jüdisches Museum Berlin
Lindenstraße 9-14
10969 Berlin
Germany
Represented by its Director, Hetty Berg 
info@jmberlin.de
T +49 (0)30 259 93 300

Data Protection Officer

We have appointed a Data Protection Officer:

dacuro GmbH
Heinrich-Hertz-Straße 11
69190 Walldorf
Germany
Email: datenschutz.extern@jmberlin.de 
Website: https://www.dacuro.de/ 

General Information on Data Processing

The protection of personal data is important to us. We process personal data exclusively in accordance with applicable data protection laws, in particular the General Data Protection Regulation (GDPR).
The application is provided as a web-based service. Personal data is processed only to the extent necessary for operating the application and providing its functionalities or where the user has expressly consented to further processing.
We do not actively collect personal details such as names, postal addresses, telephone numbers, or email addresses. However, users may voluntarily provide personal information within their conversations. Such information may be processed and stored as part of the conversation history as further described below.

Categories of Data Processed

Depending on how the application is used, the following categories of data may be processed:

• Text inputs and prompts entered by users
• Responses generated by the AI system
• Date and time of use
• Technical connection data
• IP address
• Browser type
• Operating system
• Log data (server log files)

Purpose of Processing

Personal data is processed solely for the following purposes:

• Providing the AI application
• Processing user inputs to generate responses
• Ensuring technical security and system stability
• Error analysis and misuse detection
• Publishing content shared by users
• Complying with legal obligations

User inputs are not used for advertising purposes.

Legal Bases for Processing

Processing is carried out in particular on the basis of:

• Article 6(1)(b) GDPR (performance of a contract or provision of the application service)
• Article 6(1)(f) GDPR (legitimate interests in security, error analysis, and protection against misuse)
• Article 6(1)(a) GDPR (consent), where content is published or where processing is not necessary for the use of the application

Any consent given may be withdrawn at any time with effect for the future.

Processing of Chat Content

As part of the use of the application, user inputs and AI-generated responses are processed to the extent technically necessary.
Chat histories are stored only within the browser session of the device being used and, depending on the user's browser settings, are generally deleted when the browser tab is closed.
The operator and the museum do not permanently store chat histories.
AI-generated text results are retained beyond the current session only if users choose to share them. 
Chat histories and their contents are not used to train or improve AI models.

Publication of Content

Users may voluntarily share AI-generated content and thereby publish it on a website accessible through the Jewish Museum Berlin website at: www.jmberlin.de/en/moon#results
Published content is stored for twelve months and remains publicly accessible during that period unless the user's consent is withdrawn.
The legal basis for this processing is consent pursuant to Article 6(1)(a) GDPR.
After the twelve-month retention period, the content is deleted or anonymised unless legal retention obligations or overriding legitimate interests require otherwise.
Thereafter, a selection of text examples may be published permanently on the Jewish Museum Berlin website for the purpose of documenting the exhibition and the AI application.

Use of the AI Model

A generative AI model is used to process user inputs.
The primary model used is Mistral Medium 3.5, provided by Mistral AI. If this model becomes unavailable, the application will use a locally hosted Mistral Small model as a fallback. If that model is also unavailable, the application will use Claude Haiku as a secondary fallback.
To generate responses, user inputs (prompts) and the relevant conversation context are transmitted to the AI service.
The transmitted prompts and contextual information are not stored permanently and are not used for training or improving the AI models employed.

Hosting

The application is hosted on the Microsoft Azure cloud infrastructure.
The application's backend, frontend, and database are operated in the Azure region West Europe.
The primary AI model is provided through Mistral AI's infrastructure within the European Union.
To process user inputs, the data required for generating responses is transmitted from the application to Mistral AI and processed there.
The fallback model Mistral Small is hosted on a Kubernetes cluster in Microsoft Azure West Europe.
The secondary fallback model Claude Haiku is provided by Anthropic through its global infrastructure, primarily located in the United States.

Recipients of Personal Data

In connection with the operation of the application, personal data may be disclosed to Microsoft as a technical service provider for hosting and AI-related services. Personal data is disclosed only:

• where necessary for technical operation,
• where required by law, or
• where the user has provided explicit consent.

Transfers to Third Countries

Based on the current system configuration, personal data is not transferred to countries outside the European Union (EU) or the European Economic Area (EEA), unless both Mistral models become unavailable simultaneously and the application switches to Anthropic as the second fallback provider. 
Should such a transfer occur, it will be carried out only in compliance with the requirements of Articles 44 et seq. GDPR.

Data Retention

Personal data is generally retained only for as long as necessary to fulfil the respective purpose.
Personal data, including server log files, IP addresses, system logs, and error logs, is retained for a period of two weeks and subsequently deleted unless statutory retention requirements or legitimate reasons require longer storage.
Chat histories are not stored by the operator. Storage beyond the browser session occurs only where the user expressly chooses to share or publish an AI-generated result, as described in Section 8.

Access Rights

Access to personal data is granted only to persons who require such access to perform their duties.
Administrative access to technical systems and the personal data processed within them is restricted to the responsible IT administrators.
As chat content is not stored permanently, employees do not have access to past chat histories or their contents.
The services used to operate the application run within a private Virtual Private Network (VNET) and are not directly accessible from the public internet.
Changes to systems and relevant configurations are logged.
Access permissions are limited to what is necessary for the respective task and are reviewed regularly.
A data processing agreement pursuant to Article 28 GDPR has been concluded with the service provider responsible for hosting and maintaining the application.

Data Subject Rights

Under the GDPR, users have the following rights in particular:

• Right of access (Article 15 GDPR)
• Right to rectification (Article 16 GDPR)
• Right to rectification (Article 16 GDPR)
• Right to restriction of processing (Article 18 GDPR)
• Right to data portability (Article 20 GDPR)
• Right to object (Article 21 GDPR)
• Right to withdraw consent at any time with effect for the future

Right to Lodge a Complaint

Users have the right to lodge a complaint with a supervisory authority regarding the processing of their personal data.
The competent supervisory authority is:

Federal Commissioner for Data Protection and Freedom of Information: Die Bundesbeauftragte für den Datenschutz und die Informationsfreiheit (BfDI)
Graurheindorfer Straße 153
53117 Bonn 
Germany

Automated Decision-Making

No automated decision-making within the meaning of Article 22 GDPR takes place.
The application generates AI-based content solely for assistance purposes. It does not constitute automated decision-making producing legal effects or similarly significant effects for users.

Datensicherheit

Appropriate technical and organisational measures in accordance with Article 32 GDPR are implemented to protect personal data.
These measures include, in particular, safeguards for the confidentiality, integrity, and availability of processed data, such as:

• the use of European-based AI models, 
• the absence of mandatory collection of personal information such as email addresses, 
• the non-retention of chat content, 
• and the publication of generated results only with the user's consent.

Changes to this Privacy Notice

We reserve the right to amend this Privacy Notice where necessary due to technical developments or changes in legal requirements. The version published on this website at the relevant time shall prevail.