Keeping things private: Processing sensitive text data using open-source large language models

Do you work with sensitive text data and does that make you hesitant about using AI models such as ChatGPT? Are you unsure about what really happens to your data and are you concerned about privacy and security? And most importantly: are you looking for a safer alternative to process e.g. IP- sensitive texts such as research proposals?

This course, brought to you by dr. Pieter Fivez (TEXTUA), explores the use of open-source large language models as alternatives for dealing with sensitive data in a more secure way. But does their performance match that of the likes of ChatGPT? During a lecture and a hands-on workshop, you gain insight into current large language models’ potential and implications for data safety and you tackle case studies that you can afterwards apply to your own data.

Learning outcomes

After having attended this workshop, you will...

• understand how well open-source Large Language Models perform compared to closed-source models such as ChatGPT
• have experimented with small versions of these open-source models which fit on their personal laptops.

Competences

An important part of preparing for any further professional step is becoming (more) aware of the competences you have developed and/or want to develop. In the current workshop, the following competences from the UHasselt competency overview are actively dealt with:

• academic research competences:
  • data management
  • information management
  • research methods
• intellectual competences:
  • innovativeness
  • problem solving
  • vision

For whom?

• PhD students & postdocs
• 20 places available

When and where?

• May 04, 2026 - 10:00-15:00 (lecture (10:00-12:00) & workshop (13:00-15:00))
  • Location: Campus Hasselt, building Old Prison, room OG-B1.1 GR (green room)

Preparation?

Before the training, participants should have access to:

• At least one of the following closed-source large language models:

  • Microsoft Copilot

  • ChatGPT (e.g. via chatgpt.com)

  • Gemini (e.g. via gemini.google.com)

  • Claude (e.g. via https://claude.ai)

• Open-source large language model: Ollama   

  Ollama is available through the Software Center. To install the software, your laptop must be connected to the UHasselt network. If you are not on campus, you need to connect via an EduVPN connection first. Only when your device is connected to the UHasselt network (directly or through EduVPN) will the software become visible in the Software Center. 

• Important tips:

  • Start early: Begin the installation process well in advance of the workshop to ensure everything is working correctly. 
  • UHasselt laptops: Due to updated installation permissions on UHasselt laptops, some users may experience issues. Please don’t wait until the last minute to address these, as they may take extra time to resolve.
  • For assistance, please contact the UHasselt helpdesk timely. They can help with any issues you encounter during installation. If problems persist, please inform the Doctoral Schools without delay.
    
    
• Please don't forget to bring your laptop to the session!

Registration?

• Register here from January 08 until April 06, 2026.
• As places are limited, registering does not automatically imply that you will be able to participate. You will be notified by email after the registration deadline has passed.
• Please cancel your registration at least one week in advance in case you cannot make it (cf. cancellation & no-show policy).

Acknowledged as?

• DS BSH: category 'research management & communication skills' - workshop on communication skills
• DS HLS: category 'transferable skills' - half a transferable skills course
• DS ST: category 'scientific & generic communication' - course on scientific and generic communication