Skip to content

Usage Policy

This document is a translation, only the original German document is legally binding.

Before accessing or utilizing any of our artificial intelligence services, you are required to thoroughly review and accept the following rules and usage policies.

1. Registration and Access

Access to the TUD:AI Chat Web Interface service requires a TUD ZIH user account.

Access to the API requires an API key. You may not share a key assigned to you with anyone else, unless explicitly permitted.

2. Authorized Use

Users are obligated to use the technology or services solely for authorized and lawful purposes, complying with all applicable laws, regulations, and rights of others, including national, state, provincial, local, and international laws.

3. Prohibitions

Users are prohibited from using this service to transmit, generate, or distribute content (input and output) that:

  • Constitutes child pornography or sexual abuse, including the falsification, deception, or imitation thereof
  • Is sexually explicit and used for non-educational or non-scientific purposes
  • Is discriminatory, promotes violence, hate speech, or illegal activities
  • Violates privacy laws, including the collection or distribution of personal data without consent
  • Is fraudulent, misleading, harmful, or deceptive
  • Promotes self-harm, harassment, bullying, violence, and terrorism
  • Promotes illegal activities or violate intellectual property rights and other legal and ethical boundaries in online behavior
  • Attempts to circumvent our security measures or cause actions that intentionally violate established policies
  • Could unjustifiably or adversely affect individuals, particularly with regard to sensitive or protected characteristics

Users are prohibited from engaging in the following activities:

  • Reverse engineering, decompiling, or disassembling the technology
  • Unauthorized activities such as spamming, spreading malware, or disruptive behavior that affects service quality
  • Modifying, copying, renting, selling, or distributing our service
  • Tracking or monitoring individuals without their express consent
  • Processing protected data or confidential information, unless the legal framework conditions are met

4. Termination

You may terminate your use of the service and your legal relationship with us at any time by ceasing to use the service. If you are a consumer in the EU, you have the right to withdraw from these terms within 14 days of acceptance by contacting support.

We reserve the right to suspend or terminate your access to the service or deactivate your account if you violate this usage policy, if it is necessary to comply with legal requirements, or if your use of our service poses a risk or harm to us, our users, or third parties.

We will send you a notification before deactivating your account, unless this is not possible or permitted by law. If you believe that your account has been mistakenly suspended or deactivated, you can contact support to contest this.

We reserve the right to take legal action to protect our intellectual property rights and the safety of our users. Violation of these terms or illegal activity through the use of our service may result in civil penalties, damages, administrative penalties, criminal prosecution, or other legal options.

5. Correctness of Output

The outputs generated by our services are not always unique, correct, or accurate. They may contain inaccuracies, even if they appear detailed. Users should not rely solely on these results without independently verifying their accuracy.
Furthermore, the information provided by our services may be incomplete or outdated, and some results may not align with our perspectives. Therefore, users should exercise caution and not use the services for important decisions, especially in areas such as medicine, law, finance, and other professional fields where expertise is essential. It is important to understand that AI and machine learning are constantly evolving, and although we strive to improve the accuracy and reliability of our services, users should always evaluate the accuracy of the results and ensure that they meet their specific requirements by manually verifying them before use or distribution. Furthermore, users should refrain from using results related to individuals for purposes that could have a significant impact on them, such as legal or financial decisions.
Finally, users should be aware that incomplete, inaccurate, or offensive results may occur that do not reflect the views of ScaDS.AI or related parties.

6. Limitations of Liability

General limitation of liability: ScaDS.AI accepts no liability for claims for damages by users based on the use of the service. The limitation of liability set out here and explained in more detail in the following sections is based on the fact that ScaDS.AI exclusively provides a platform for the use of language models. ScaDS.AI cannot provide any technical measures on this platform to intervene in the responses generated by the language models in such a way that damage due to improper use by users can be ruled out. Therefore, users remain fully liable for claims for damages resulting from improper use of this platform. This does not apply to claims based on injury to life, limb, or health, gross negligence, or intentional or grossly negligent breach of duty. Likewise, the breach of cardinal obligations is excluded from the fundamental exclusion of liability.

Copyright: Users of the service are fully and solely responsible for observing and complying with the applicable provisions of copyright law. ScaDS.AI expressly points out to users that the language models provided have been trained by third parties and that ScaDS.AI has no declaration restricting the materials used to free licenses. Consequently, ScaDS.AI cannot rule out the possibility that the language models provided have been trained using copyright-protected content. Responses provided by the language models to users may therefore contain copyright-protected content. ScaDS.AI explicitly advises users that direct reuse of the responses received is not recommended. The responsibility for checking the copyright in such cases lies solely with the users. ScaDS.AI accepts no liability for any claims for damages arising from copyright infringements.

Confidential information: We cannot accept any liability for the loss/publication of data provided by users in their inquiries. This does not apply in cases of gross negligence or cardinal obligations.

Patent law: Users of the service are solely responsible for observing and complying with the applicable provisions of patent law. Responses provided by the language models may contain patent-protected content in their conceptual design. ScaDS.AI explicitly advises users that direct reuse of ideas and concepts conveyed in the responses received is not recommended. The responsibility for checking patent protection in such cases lies solely with the users. ScaDS.AI accepts no liability for any claims for damages arising from patent infringements.

Misinformation: ScaDS.AI advises users of the service that it is an intrinsic characteristic of the language models provided to freely invent content — this is referred to as “hallucination” in language models. The information contained in the responses may be outdated, fictitious, inappropriate, taken out of context, or incorrect. This does not constitute a malfunction of the platform provided, as it is technically to be expected from the language model tool. The independent and critical review of the information received is the sole responsibility of the user. ScaDS.AI accepts no liability for the information contained in the responses of the language models. For more information, see “5. Correctness of Output”.

When using the RAG system, users bear sole responsibility for the legality of the documents they upload, store, and index in collections. ScaDS.AI accepts no liability whatsoever for the documents uploaded by users to the RAG system.

7. Feedback

We value your feedback on our services and products and encourage you to share your thoughts to help us improve. By providing feedback, you understand that we may disclose, publish, exploit, or use it to improve our offerings without owing you any compensation. We reserve the right to use feedback for any purpose without restriction by confidentiality obligations, regardless of whether it is marked as confidential or not.

8. Data Security

See Chat Web Interface and API.

9. Final clauses

Even if individual points of this usage policy are legally invalid, the remaining parts shall remain binding and effective. The invalid points shall be replaced by the statutory provisions, if any. However, if this would constitute an unreasonable hardship for one of the contracting parties, the contract shall become invalid in its entirety.