Disclaimer UU AI-Chat

For Dutch click here.

UU AI-Chat is an AI environment developed by Utrecht University that uses AI models from providers including OpenAI, Mistral and SURF. The tool uses large language models (LLMs), a form of generative AI that can generate text and images based on user input. These models have been trained on large amounts of data, which they use to generate appropriate responses.

Whilst this technology offers many possibilities, it is important to be aware of its limitations. UU AI-Chat does not have knowledge or understanding of its own, and the information it generates may not always be complete, accurate or up to date. Responsible and critical use is therefore essential: use AI as a tool, not as a substitute for your own judgement or reliable sources.

Prohibited uses

It is not permitted to:

present UU AI-Chat as a real person,
use UU AI-Chat to make predictions about people’s behaviour,
use UU AI-Chat to recognise the emotions of UU staff and students,
use UU AI-Chat to recognise or identify people in images,
use UU AI-Chat to determine whether someone is granted access to a degree programme, module or course, or to determine which level of education would be suitable for someone,
use UU AI-Chat to assess student work, regardless of whether a grade is awarded,
use UU AI-Chat to detect plagiarism or fraud among students,
use UU AI-Chat to assess job applicants or the documents they have submitted, and
use UU AI-Chat in processes relating to an employee’s employment relationship.

Check the output

AI can make mistakes. Output generated by UU AI-Chat may be inaccurate, incomplete or outdated. The tool may also generate information, sources or references that do not exist or do not correspond to the original source. You

should therefore always check the output carefully, especially if you are using it for your work. You remain responsible for assessing the output and for how you use it.

Bias

AI models are trained on large amounts of data. This data may contain biases, stereotypes and unbalanced perspectives. This may be reflected in the output of UU AI-Chat. You should therefore always assess AI-generated output critically, particularly when it relates to individuals or groups.

Privacy and sensitive information

UU AI-Chat applies strict privacy measures when processing data. Although UU AI-Chat is suitable for processing sensitive information, it is not permitted to enter special categories of personal data (information for students, information for employees), such as political opinions or health data, unless you have the appropriate privacy documentation (information for employees) in place.

Under no circumstances is it allowed to enter medical patient information into UU AI-Chat.

For information on how UU AI-Chat processes your data, please see the Privacy Statement.

Not a substitute for professional advice

UU AI-Chat is a tool and not a substitute for professional expertise or human judgement. Do not regard its output as legal, medical, financial or any other form of professional advice. Consult an expert when professional expertise is required.

Copyright

AI models may have been trained on copyright-protected material. Generated output may also resemble existing works. You should therefore not automatically assume that all AI-generated output may be freely used, published or distributed. Be mindful of copyright, licences and other third-party rights.

Transparent and responsible use

Utrecht University has rules on the responsible and transparent use of generative AI. These are set out in the University’s AI policy. Make sure you are familiar with this policy and the rules that apply to your use of AI.

Always use UU AI-Chat in accordance with the AI policy, applicable laws and regulations, and other relevant guidelines. Additional rules may apply within your faculty, degree programme, course or organisational unit.

Tools

Privacy Statements