Strategic Foresight in Uncertain Times

Antonia Witt at the CCEW Symposium in Berlin

In October, inter­national experts from research, think tanks and ministries concerned spent two days discussing the limits and challenges of conflict early warning, forecasting and foresight. How will crisis early warning systems evolve in an era of inter­state conflict, geo­politics and hybrid threats? What impact will the use of AI and large language models have? What new approaches and methods are being developed for conflict fore­casting and strategic fore­sight?

Six panels and three seminars contributed to an inter­disciplinary ex­change on these various aspects of effective early detection and foresight, and enabled discussions between researchers and decision-makers.

Antonia Witt is one of the project leads of JOSHUA – AI-based Scenario Development, a research project at PRIF in cooperation with the chair “Science and Technology for Peace and Security” and the “Artificial Intelli­gence and Machine Learning (AIML) Lab” at TU Darmstadt, as well as the German Research Center for Artificial Intelli­gence (DFKI). At the CCEW Symposium, she led a work­shop on the possibilities of AI-supported fore­sight and scenario develop­ment and, together with Dominik Hintersdorf and Marlon May (both DFKI), presented the project’s research and develop­ment results to date. The system developed in JOSHUA uses lan­guage models and agent-based modelling to simulate conflict dynamics. This makes it possible to inter­actively explore different “what-if” scenarios for inter­state and intra­state conflicts, test strategies, and reflect on escalation and de-escalation dynamics. In her presentation, Antonia Witt emphasized that, beyond processing large volumes of data, artificial intelligence can be used in strategic fore­sight as a “creative partner” to broaden the imaginative horizon for multiple future path­ways. At the same time, this requires responsible, trans­disciplinary and trans­parent develop­ment and use of AI-based systems, with human-machine inter­action always subject to critical and ethical re­flection.

The CCEW Symposium is organized annually by the Center for Intelli­gence and Security (CISS) at the University of the Bundes­wehr Munich and took place this year at the Federal Academy for Security Policy in Berlin. It is supported by the Federal Ministry of Defence and the Federal Foreign Office.