Integreat workshop: AI for justice

Researchers and representatives from civil society seek to investigate the impact of generative AI in relation to justice and war crimes.

Time and place: –, University of Oslo, Blindern, Georg Svedrups hus, LINKEN meeting room
Bildet kan inneholde: elektronisk enhet, font, multimedia, skjermdump, programvare.

Description

The workshop gathers representatives of academia, civil society and expert communities working with Artificial Intelligence and/or war crimes databases and archives. Workshop participants will learn about the recent advances in generative AI and the new possibilities offered by these AI tools for generating and extracting information from text, audio, images and videos. The workshop is the occasion to discuss the specificities of war crime related data and test AI models on these data.

During the workshop, participants will test AI models in practice helped by the teaching team from Integreat. Mixed group with participants from diverse backgrounds will work on applying AI to archive material (either online data, data provided by NGOs or their own data).

Expected outcome

Through individual tasks and teamwork in mixed interdisciplinary groups, the participants will network, share their professional experiences and build a small AI-aware community. This community will specialise in using Generative AI to advance the effective systematisation, preservation, presentation and analysis of war crimes archives. Participants will get access to code and tools shown during the workshop and can use them to make their own archives/databases AI-enabled. The code will be available on the AI for Justice Github repository.

Programme

Day 1, 11 June

09.00 Welcome & Introduction by the organisers
09.10
Short overview of Generative AI and potential applications to War Crimes Archiving

Associate Professor Benjamin Ricaud, UiT The Arctic University of Norway
09.30
Role of International Institutions in Documentation of Core International Crimes

International Court NGO Representative (to be confirmed)
09.50 Coffee break
10.00
Ukraine War Crimes Trial Database

Associate Professor Megumi Ochi, Ritsumeikan University, Japan
10.20
Short presentations of own databases and participants’ experiences with digital archiving and Artificial Intelligence (AI)
11.00
Digital Evidence and OSINT in Documenting War Crimes in Ukraine

Maryna Slobodyanuk, Coordinator for Training and Strategic Projects, NGO Truth Hounds, Ukraine
11.20 Lunch
12.30
Retrieval Augmented Generation and chatbots, new paradigm for archiving and retrieving documents

PhD Fellow Daniel Kaiser, UiT The Arctic University of Norway
12.50
Practice part 1: Retrieval Augmented Generation and chatbots

Goal: build a prototype to archive data, understand how it works and the limits. Reflect on the limitations. Setting up a RAG system (with or without coding skills), step by step tutorial with participants. Application to different data, participants work on some data they chose. Open discussion about the possibilities and limits.
14.00
The Role and Legal Framework of OSINT in the Collection and Analysis of Digital Evidence in the Ukrainian case

Marie Curie Fellow, Associate Professor Kateryna Latysh, Yaroslav Mudryi National Law University, Ukraine; Senior Lecturer, Vilnius University, Lithuania
14.20 Coffee break
14.40
Practice part 2: AI for video, images and audio

Goal: get to know AI models which extract information from video, audio and images. Practice with open-source models from Huggingface.co, WHISPER, CLIP and others, convert speech to text, image to text and produce video summaries. Application to different data, participants work on some data they chose. Open discussion about the possibilities and limits.
16.00 End of the first day
17.00 Pizzas

Day 2, 12 June

09.00
MSCA Project ‘Hybrid Cyber Warfare and Common Security in Europe’, Open-Access Book on OSINT and digital investigation of core international crimes

Marie Curie Postdoctoral Fellow Artem Galushko, UiT The Arctic University of Norway
09.20
Managing databases and digital archives: challenges and opportunities in Ukraine

Valentyna Soloviova, Analytical Department Manager, NGO Truth Hounds, Ukraine
09.40 Coffee break
10.00
Practice part 3: Understanding the AI representation of data

Goal: understand how an AI model represents the data and organizes it. Understand the imperfect representation of AI models and the risks of relying too much on them.
Activities: Tutorial on data visualization, t-SNE, UMAP, Graphs with Gephi and application to AI models. Improving the RAG model, using what we have learned about AI model representations.
11.30 Lunch
12.30
Practice part 4: Further developing AI for archiving

Open session for the participants’ projects, combining the AI seen in part 1, 2, and 3. Participants work on real-world archive data in teams, developing further a RAG system or some data analysis. The teaching team assists the teams in their projects.
14.00 Break & small walk around the campus
15.00 Group Presentations and Feedback from Trainers/Group Leaders
15.45 Coffee break
16.00 Final Discussion and Closing

Registration

Registration to the workshop is required.

Registration form (30 May)

Participants should have either
  • some programming background,
  • experience with databases,
  • or a knowledge of or an interest in the process of war crime archiving and investigations.

Practical information

  • Lunch will be served to registered attendees.
  • Streaming will be available for attendees joining online.

Points of contact

Organiser

Integreat, the UiT hub
Published May 13, 2025 11:33 AM - Last modified May 13, 2025 11:53 AM
E-mail this page