Working with AI
AI offers opportunities to make your work easier. However, there is also a downside to AI that may have consequences for your privacy and that of your colleagues. On this web page, you will find more information about the responsible use of AI within Leiden University.
AI use: the golden rules
AI offers many opportunities, but also entails risks. Therefore, remain critical, check the results, and do not simply accept information at face value. Technology provides support, but the responsibility for the generated output always lies with the user.
When working with AI, it is important to handle data, information, and technology consciously and carefully. In this context, take the following into account:
You can use the University’s tool picker to see which tools you are allowed to use. This also applies to AI. Find your AI model in the tool picker and check whether you are allowed to use it. Green means ‘permitted’, orange means ‘permitted in exceptional cases’, and red means ‘not permitted'. If there is no green tool available (or not yet), you may use an orange tool until further notice. In that case, please observe all the rules above.
Personal data are data that can be used to identify someone directly or indirectly. This includes names, unique job titles, photographs, or employee numbers. Never use these data with an AI model.
If you want to share SharePoint, Teams, OneDrive or Outlook documents with an AI tool, always check the confidentiality label that is automatically assigned to the document in question, or that you have manually assigned to it. Documents can be labelled Public, Internal, Confidential, or Secret. Only documents labelled ‘Public’ may be shared with an AI tool.
You can find more information about sensitivity labels here.
Information that is not recorded in a document but falls into the ‘Internal’, ‘Confidential’, or ‘Secret’ categories may not be shared with an AI tool. Think of information that has been shared with you verbally. If you do share such information, this constitutes a data breach.
Check whether the information generated by AI is factually correct and appropriate for its use. For example, check with your colleague or an expert within your department.
Risks of AI
AI offers many opportunities, but it also entails risks. So keep thinking carefully, keep checking, and don’t blindly accept content. Technology is a tool, and it is not responsible for the output used.
FAQ
Yes, you may use AI for your work. AI offers many options that enhance the education we provide or that can simplify our work. However, it is important to use AI responsibly, and to always observe the five golden rules.
If there is no green tool available (or not yet), you may use an orange tool until further notice. Remember to observe the golden rules.
At present there are no AI tools that meet all our security and privacy requirements. The security of your data and your privacy cannot be guaranteed. We understand that there is a need to use AI tools, so in the meantime, you may use Copilot.
No, ChatGPT is not supported by Leiden University. The security of your data and your privacy are not guaranteed. Anything you enter into ChatGPT becomes part of the data set used to train the AI model. Leiden University no longer has any control over it.
In the tool picker, you can see which other tools are not supported. If you would still like to use an AI tool, then Copilot is the best choice for now. Only share public information with Copilot.
We are glad to hear that you are using the tool picker. If you would like to have an AI tool included in the tool picker, please contact your ICT contact person. Your suggested AI tool will be assessed. You will receive feedback about an alternative or the security or privacy checks that still need to be carried out.
We are currently working hard on an AI policy. Once this policy has been finalised, it will be shared internally.
More information and questions
If you would like to know more about AI in education, see the information on the AI in Education page.
If you have any questions that are not answered above, please contact the Security Office for questions related to information security at security@bb.leidenuniv.nl, or the Privacy Office for questions related to information protection and privacy at privacy@bb.leidenuniv.nl