Building Stronger Futures Requires Better Questions

Reflections from ISCRAM 2026 in The Hague

From May 31 to June 3, researchers and practitioners from around the world gathered at Leiden University in The Hague for ISCRAM 2026, the annual conference of the Information Systems for Crisis Response and Management (ISCRAM) community.

The conference theme, Building Stronger Futures, focused on the interconnected challenges of natural hazards, cybersecurity, AI/technology, critical infrastructure, climate change, and public health. Underpinning many of the conversations was a "whole-of-society" approach to crisis management: the recognition that resilience emerges not only from institutions and technologies, but from the interactions among communities, organizations, and individuals.

Over four days, I attended a simulation tutorial and sessions on simulation design for emergency planners and universal design. As I attended the different sessions, I found myself returning to the same questions:

• How do we build models that are useful without oversimplifying reality?

• How do we elicit knowledge from people with different forms of expertise?

• How do we faithfully and ethically consider intersectionality of social factors/identities in the parameters and objectives of the models and tools we build?

These are fundamentally interdisciplinary challenges. They sit at the intersection of technology, social science, design, policy, and lived experience. They are also questions that have been at the heart of Storymodelers' work for more than a decade.

One of the highlights of the conference was the launch of a new ISCRAM track: Gender Perspectives, Diversity, Equity and Inclusion in Emergency Management.

The innaugural track for ISCRAM was organized by Storymodelers collaborators Hanne Haaland and Hege Wallevik from the University of Agder, together with their colleague Terje Gjøsæter at UiA's Centre for Integrated Emergency Management. Although this was the first year for the track, the collaboration behind it has much deeper roots. Hanne and Hege have been part of the broader Storymodelers community for nearly 10 years, exploring questions of participation, narratives, social complexity, and knowledge elicitation.

The interest in these conversations was clear from the start; to our surprise, 17 people attended and nearly filled the room and engaged in a lively interdisciplinary conversation after the presentations across computer science, engineering, and feminist studies.

That enthusiasm felt fitting. Interdisciplinary work is often discussed in terms of methods and tools. Less attention is paid to the trust required to develop them. Eliciting models from communities, practitioners, policymakers, and researchers is not something that happens in a single meeting or workshop. It takes years of collaboration, shared language, and mutual understanding about what counts as knowledge and how different perspectives can be brought together. These are lessons Storymodelers have learned in spades and continue to contemplate as we do our research. In many ways, the new track represented the culmination of those conversations.

The session opened with a presentation by Hanne, Hege, and Terje examining how gender has been addressed in ISCRAM research over the past two decades. Their systematic review found growing attention to gender and diversity within the field, while also highlighting opportunities for more explicit intersectional approaches.

A recurring theme was that diversity is often treated as a set of categories rather than as overlapping experiences and social identities (e.g. intersectionality) that shape both vulnerability and resilience during emergencies and protracted crises.

That insight carried into the second presentation of the track, delivered by Storymodelers researcher Guljannat Huseynli and me. We explored a growing challenge in humanitarian response: how AI systems balance efficiency and equity.

An active participant activity to explore how they conceptualize humanitarian AI systems - conducted live in our presentation.

Humanitarian organizations increasingly use AI to forecast crises, assess damage, allocate resources, and manage information. But every system simplifies reality. We require an analytical abstraction for efficiency in models. To function, organizations and modelers must decide what counts, what can be measured, and what will be optimized.

Intersectionality reminds us that people's experiences are shaped by multiple, interacting factors—such as gender, disability, age, displacement, and socioeconomic position. The same layers of identity and social relationships that create resilience can also create vulnerabilities that become difficult to see when systems reduce people to categories, risk scores, or administrative units.

Our proposed governance framework encourages organizations to look beyond technical performance and ask broader questions - earlier questions in the design process - about representation, participation, and accountability. Which experiences become visible in the data? Which may be excluded by design? Given the black box of AI, we cannot wait for outputs of AI-decisionmaking or management tools to ask these questions, because it can become administratively difficult or impossible to correct misclassifications, blindspots, or miscalculations.

An overview of our draft framework for an Intersectional Governance Audit of AI-informed tools and models for humanitarian or disaster management and response.

These are not purely technical questions. They require ongoing engagement with communities, practitioners, and decision-makers—echoing many of the knowledge-elicitation challenges discussed throughout the conference.

The conversation continued well beyond the presentations themselves, as participants explored how intersectional thinking might be incorporated more explicitly into emergency management research, technology design, and organizational practice.

As the conference came to a close, I found myself connecting our session back to many of the others I had attended throughout the week.

Whether the discussion centered on simulation design, accessibility, emergency planning, or artificial intelligence, participants were wrestling with a common challenge: translating complex human realities into models, systems, and decisions.

• Every model simplifies.

• Every dataset leaves something out.

• Every design choice makes certain experiences visible while obscuring others.

The challenge is not eliminating simplification through abstraction—that is impossible. The challenge is becoming more deliberate about what is simplified, why, and with what consequences. It's about making sure that those choices are deliberate, eyes wide open, so that outcomes of models and systems reflect the organization's intentions, expectations, and values in their outputs.

That is ultimately what intersectionality contributes to disaster and emergency management research and practice. It's why this track is an important new addition to ISCRAM. It reminds us that resilience and vulnerability emerge through layers of identity, relationships, institutions, and power structures that interact in ways our models do not always capture. It is important to turn an intentional, analytical eye towards these considerations to improve humanitarian and disaster response.

Building stronger futures requires better tools.

But it also requires better questions.

Because sometimes the most important question is not what a model can see—it is what it cannot.