Terms of Use

1. Definitions

1.1 “AI Input” means all inputs transmitted to the Services, including (i) text inputs (e.g. prompts), files, images and structured data, (ii) system instructions and configuration parameters, (iii) contextual information from previous interactions (conversation history) and from connected data sources (e.g. Retrieval Augmented Generation).

1.2 “AI Output” means any output generated by a Service using AI functionality, regardless of whether it is displayed directly to the User, forwarded to a downstream system or logged.

2. Intended Use

The User may use the Service exclusively in accordance with the intended purpose described in the Documentation. Any other or further use shall be deemed not to be as intended. The Services may pose risks if used other than as intended and/or in a different manner.

3. Restrictions on Use

3.1 Use in High-Risk Environments

The Services are not intended to be used in or in connection with the operation of facilities, products or systems whose malfunction could result in a risk to life or limb or in damage to property or the environment.

3.2 AI Input

The following data must not be used as AI Input: sensitive data, regardless of whether it constitutes personal data, e.g. trade and business secrets, data subject to specific statutory confidentiality obligations, and data that the User otherwise classifies as sensitive or strictly confidential information.

3.3 Prohibited Use

3.3.1 The Services must not be used for prohibited practices within the meaning of Art. 5 of the AI Act (Regulation (EU) 2024/1689), in particular:

• deploying subliminal techniques beyond a person’s consciousness or purposefully manipulative or deceptive techniques;
• generating or using media with sexual or sexualised content;
• exploiting the vulnerabilities or need for protection of a natural person or a specific group of persons due to their age, disability or a specific social or economic situation;
• evaluating or classifying natural persons or groups of persons over a certain period of time based on their social behaviour or known, inferred or predicted personal characteristics or personality traits;
• carrying out risk assessments in order to assess or predict the risk of a natural person committing a criminal offence;
• creating or expanding facial recognition databases;
• inferring emotions of a natural person in the workplace and in educational institutions;
• biometric categorisation of natural persons; and/or
• using real-time remote biometric identification systems in publicly accessible spaces for the purposes of law enforcement.

3.3.2 The User must not use, alter, modify or further develop any of the Services in a manner that results in the Service becoming a “high-risk AI system” within the meaning of the AI Act (Regulation (EU) 2024/1689).

3.3.3 When using the Services, the User must comply with all applicable statutory provisions.

3.3.4 The User must not use the Services, or enable or permit others to use them, in a manner that results in or aims at any of the following:

• infringement of the rights of another person, e.g. the right to life and physical integrity, personality rights, data protection rights, rights to one’s own image, industrial property rights and copyrights;
• direct or indirect discrimination against a person on the grounds of their ethnic origin, sex, age or disability, religion or belief, or sexual identity;
• creating, distributing, using or otherwise promoting pornographic, child or youth pornographic, violence-glorifying or terrorist content or activities;
• deception, fraud or misleading conduct, including through activities related to disinformation, plagiarism, spam, scams, deep fakes, phishing or malware, and the promotion of misinformation;
• breaching the security, confidentiality, integrity or authenticity of information and communication systems of ZEP or third parties, in particular of the Services themselves.

3.4 Circumvention of Security Measures

The User is prohibited from altering or circumventing security measures in the Services, in particular those that control, filter or restrict AI Input or AI Output.

3.5 Extraction of Results

The automated extraction of AI Output and other results from the Services is prohibited.

4. Responsibility of the User

4.1 The selection, use and results of the Services are the sole responsibility of the User.

4.2 The Services are not a knowledge base, research tool or reference work. AI Output is a mathematical calculation of probabilities within the meaning of the AI model used, which is largely based on its training data. The User understands and accepts that, given the probabilistic nature of machine learning, the use of artificial intelligence technologies may in some situations lead to incorrect results that do not accurately reflect real persons, places, facts or the desired AI Output. It is therefore the sole responsibility of the User to assess the suitability of the AI Output on a case-by-case basis, in particular through human review of the AI Output.

4.3 There is no guarantee that the AI Output is unique compared to (i) the AI Output of other users and (ii) pre-existing works. The User must therefore ensure that a sufficient distance from pre-existing works is maintained when using the AI Output.

4.4 The User is responsible for complying with any statutory labelling and transparency obligations.

5. Third-Party Terms of Use

The User shall use the Services in accordance with the current version of any applicable rules, requirements or policies of the third-party providers used to provide the Services. At present, these include the following: https://aws.amazon.com/de/legal/bedrock/third-party-models/