Part 1: Before you start
About this manual
This manual for UvA AI Chat is designed as a reference guide where you can look up information about the various features and settings of UvA AI Chat. If you're using UvA AI Chat for the first time - or if you don't have much experience with generative AI - we recommend starting with the "Getting Started with the UvA AI Chat" e-learning. This e-learning module walks you through all the basic features step by step and shows examples of how you can use the tool. Once you're familiar with UvA AI Chat, you can use this manual as a reference whenever you need it. You can also use the Technical Support Assistant within UvA AI Chat, which is based on this guide.
Because AI is evolving rapidly, this manual will also be regularly updated. The latest features and applications of UvA AI Chat will be included in future versions.
For questions about this guide and UvA AI Chat, please contact ai-tlc@uva.nl.
This manual is developed by TLC Central (see tlc.uva.nl)
Version 1.3 | Published 1st of April 2026
1.1 Responsible use
It's essential to understand that UvA AI Chat, like any other AI tool, is not infallible. The model can present mistakes, inaccuracies, biases, or incorrect information as if it were fact. Never rely blindly on the generated output. Always verify the results carefully and apply critical judgement each time you use it. Also keep in mind that using AI consumes both energy and water.
Practical example of responsible use
Suppose you ask the AI to summarize a scientific article you've uploaded. The AI produces a concise summary.
- Irresponsible use: You copy the summary directly and use it in your own work without any verification.
- Responsible use: You read the summary generated by the AI and compare it carefully to the original article. You check whether the main arguments, methodology, and conclusions are represented correctly. You use the AI output as a tool to help you grasp the article faster, but you base your own work on the actual source text.
Part of using generative AI responsibly is being aware of the possible downsides. When using AI, you should be aware of the risks of cognitive offloading. Cognitive offloading refers to the process of transferring cognitive processes to an external tool, such as an AI. Especially in an educational setting it is important to be aware that continuous reliance on AI for a particular task might hinder you from developing knowledge and skills that would benefit you in the long run. Research shows that using AI frequently but unreflectively can lessen critical thinking abilities.
To use AI effectively and responsibly, we need to develop our AI literacy. Among other things, this means understanding how to use AI tools well, like asking the right questions, and carefully checking what the AI produces. By developing our AI literacy we can ensure that we make use of AI in such a way that it supports our work, learning, and research in a valuable manner, while mitigating the risks the use of AI poses.
1.2 Your first step: logging in
Open AI Chat in your webbrowser. Enter your UvA login details - this means your email address and password linked to your UvA account. For additional security, two-factor authentication (2FA) is required; you'll need to enter a verification code sent to your mobile device. After successful authentication, you'll have access to the UvA AI Chat dashboard and can use all its features.
1.3 The Technical Support Assistant in UvA AI Chat
What is the Technical Support Assistant?
The Technical Support Assistant is a standard, pre-configured chatbot (persona) in UvA AI Chat. Unlike the general AI, which draws on broad training data and can answer almost any question, the Technical Support Assistant operates within a strictly defined knowledge boundary: it only uses the official UvA AI Chat manual and the associated privacy documentation as its knowledge base.
This means:
- It only answers questions that can be answered based on the manual and privacy docs.
- It provides reliable, policy-consistent answers about UvA AI Chat's features and data handling.
When should you use the Technical Support Assistant?
You can use it to get accurate, helpful answers to all your questions about UvA AI Chat, whether they relate to its features, functionalities, capabilities, or anything else you may want to know about the platform. It is especially useful in the following situations:
| Situation | Why the Technical Support Assistant helps |
|---|---|
| You're unsure how a feature works | It explains features based on the official manual, not guesswork |
| You have a privacy question about uploaded documents | It answers strictly from the privacy policy |
| You want to know which AI models are available | It provides the up-to-date model overview from the manual |
| You're new to UvA AI Chat | It gives reliable information for new users |
| You need to verify what UvA AI Chat can and cannot do | It won't overstate or invent capabilities |
We do not recommend using the Technical Support Assistant for general academic tasks, writing assistance, or anything outside the scope of UvA AI Chat's own documentation - it will either decline or be unable to help with those.
How to activate the Technical Support Assistant?
- Click the "?" icon in the left sidebar of UvA AI Chat.
- Select Talk to Technical Support Assistant.
- Start a new conversation: the Technical Support Assistant is now active.
Practical use cases
Use case 1: understanding a feature
You ask: "How does the prompt library work?"
Support assistant answers by explaining what the prompt library is, where to find it, and how it works.
Use case 2: understanding personas themselves
You ask: "What is a persona and how do I create one?"
Support assistant answers by walking you through the 7-step persona creation process from section 3.2 of the manual, including setting a title, description, model, conversation style, behavior instructions, and uploading knowledge files.
1.4 Mobile interface
The UvA AI Chat can also be used via the browser on your phone: no app installation required. Simply open the website on your phone and you will see an interface that is virtually identical to the desktop version.
The navigation menu
The main difference from the desktop version is the navigation menu. On a computer, this panel is expanded by default on the left side of the screen. On your phone, the menu is hidden by default to save space.
You can open the menu using the arrow in the top-left corner of the screen. The menu will slide open and display the same options as on the desktop, shown with icons:
| Icon | Function |
|---|---|
| Speech bubble | General chat |
| Mask | Personas |
| Folder | Projects |
| Two people | Groups |
| Booklet | Prompts |
| Puzzle piece | Extensions |
At the bottom of the menu you will find:
- ? — Additional materials, such as the manual and E-learning
- Person icon — Settings and account options
Tap on the arrow in the top left of the menu again to close it.
Other features
All other features work the same as in the desktop version. Refer to the relevant sections in this manual for more information on the individual options.