Panel I: Novel Approaches to Forecasting Protest

Chair: Maxine Leis, Uppsala University

Recent years have seen a substantial increase in academic research on forecasting events besides armed conflict. This panel presents and discusses novel approaches to predicting protest movements. In terms of methodology, the panel investigates ways of improving transparency in detecting protest from images and methodological innovations in understanding how deaths caused by state repression affect protest cycles. Furthermore, the panel discusses the proposal of a new data source to measure levels of contention.


Panel II: Disentangling Conflict Prediction

Chair: Sonja Häffner, University of the Bundeswehr Munich

The introduction of new datasets and recent advancements in computing power, accompanied by methodological innovations, has improved our ability to forecast conflict. Nevertheless, reliably predicting conflicts remains a persistent challenge. We will unravel some of the numerous complexities in this session by exploring alternative approaches. On the one hand, we will investigate the potential of prediction markets and an ontology derived from systems theory to improve conflict prediction. On the other hand, the properties of the evaluation metric, TADDA score, will be assessed.


Panel III: Anticipating the Behaviour of Non-State-Actors

Chair: Christoph Dworschak, University of York

Civil conflicts are multi-layered and complex environments. Various non-state actors compete with each other and the government for power and territorial control. Interactions between violent and non-violent actors further complicate these dynamics. In this panel, we discuss actor-centric approaches to predict the severity of conflict. In addition, we engage in a debate on how to detect and prevent actor-specific behaviour such as illicit commodity trafficking.


Panel IV: Regional Approaches in Crisis Early Warning

Chair: Melanie Sauter, University of Oslo

Regional approaches to crisis early warning have become an increasingly important trend in recent years. This approach recognizes that many crises often occur within specific geographic contexts. This panel discusses regional approaches to predict distinct phenomenas including internet shutdowns in Africa, targeting of civilians in the Russian war against Ukraine, and the testing of hypotheses on Biased Conflict Fatality Estimates in different regional conflicts. 


Panel V: Translating Data into Action

Chair: Hannah Frank, Trinity College Dublin

Despite significant advances in conflict prediction and crisis early warning, anticipatory and pre-emptive action on behalf of practitioners and policymakers are hindered by organizational and practical constraints. This panel aims to reflect on ways to translate crisis early warning data into action that may prevent conflict escalation and help minimize impacts, both through the lenses of academics and practitioners. On the one hand, this panel investigates features of an ideal Early Warning System. On the other hand, it discsusses ways to improve the conveying and receiving of politically or organisationally inconvenient warnings within intra- and supranational organizations.


Policy Panel I: From Data to Action: Broadening the Crisis Assessment

Moderator: Nanténé Coulibaly, Senior Early Warning Analyst, ECOWAS

The international community finds itself confronted by multiple, complex, and often interconnected crises. They range from ongoing conflicts in Ukraine, Somalia, or Yemen to dealing with far-reaching repercussions of climate change, economic downturns, and public health emergencies. In the face of multifaceted crises, risk assessment is crucial to every decision-making and planning process. The inherent goal is quantifying risks, dealing with uncertainties, and acting preventively. Data-driven forecasts play an essential role in dealing with such challenges. Especially AI-supported early crisis detection has made considerable advances in recent years. This panel will discuss the advantages and challenges of early warning tools.


Policy Panel II: Reach for the Stars: Outer Space in Crisis Early Warning

Moderator: Lisa Storck, European Parliament

The unrestricted use and independent access to space are essential to a state’s functioning in the civil and military spheres. However, outer space is an increasingly contested domain, playing a significant role in armed conflict and as a source of major power rivalries. Adversary actors are increasing their counter-space capabilities, while the number of non-state actors in outer space is growing exponentially. At the same time, space technology can also improve the anticipation and early detection of crises. How can Germany and Europe effectively protect themselves against threats such as the spread of space debris, the deliberate manipulation and destruction of satellites, and the vulnerability of space systems to cyber-attacks? What measures, possibly systemic changes, are necessary? How can space technology be used in crisis early warning? Which capabilities are already available today? Which will be needed in the future? Which measures and instruments are necessary to effectively integrate space aspects into the national crisis early warning process? Which role can civilian and commercial actors play in space-based crisis early warning? This panel aims to reflect on these questions from different disciplinary angles critically.