Disclaimer

Tilly is an AI chatbot platform developed for Tilburg University to support education, research, and academic services. The platform uses Large Language Models (LLMs), a form of generative AI capable of producing text based on user input.

Tilly is trained on large collections of data and can provide contextual answers to questions or comments. While this technology offers valuable opportunities for studying, teaching, and research support, it is important to recognize its limitations.

Tilly does not possess its own knowledge, understanding, or consciousness, and the information it generates may not always be complete, correct, or up to date. Users should therefore apply critical judgement and verify important information. The chatbot is a support tool, not a replacement for human expertise, academic sources, or independent thinking.

Responsible Use – Prohibited and High-Risk Tasks

Use Tilly only for legitimate university-related purposes. The way generative AI is used determines its level of risk. Certain uses are not allowed or are unsuitable at this time. These include, for example:

  • Making decisions about admission to programmes
  • Evaluating or grading academic performance
  • Determining academic level or student progress

Privacy and Data Protection

Tilly must not be used to share personal or sensitive data, including health information, political beliefs, religious details, or other confidential personal material. Permitted uses include:

  • Academic conversations
  • Anonymised data
  • Educational materials without sensitive personal information

Only basic login details are being stored for access control (e.g., name, email address, and role). Users can delete their previous conversations and uploaded content permanently. For more information about data privacy, please check our document "data privacy".

AI Literacy and Critical Thinking

Responsible AI use begins with understanding how the system works. Tilburg University encourages all users to learn about the opportunities and risks of generative AI, such as bias, accuracy limitations, and environmental impact. Therefore we have created e-modules for students and teachers/employees, which can be found on student.tilburg.ai and learn.tilburg.ai.

Reliability and Verification

Tilly predicts responses based on data patterns rather than expressing factual certainty. Answers may appear convincing even when they are incorrect. Users remain responsible for the interpretation and use of output. Critical evaluation and academic verification are essential.