From signals to systems: the emerging practice of anticipatory action for conflict-induced displacement

From signals to systems: the emerging practice of anticipatory action for conflict-induced displacement

Over the past decade, anticipatory action has evolved from a niche innovation into an increasingly mainstream humanitarian approach. Yet, despite conflict and violence remaining among the principal drivers of humanitarian needs and forced displacement globally, anticipatory action for fragility, conflict and violence (FCV) remains underdeveloped. Of the 262 active frameworks for anticipatory action in 2025, only a small fraction focused on conflict and violence, while displacement remained a marginal target risk. This gap is hard to justify: more than 113 million people were forcibly displaced at the end of 2025, and a further 4.2 million people could be displaced by the end of 2027.

The issue is not a lack of information; humanitarian systems already generate substantial early warnings and risk analysis about conflict and displacement, and across many contexts. While the precise trajectory of conflict is often uncertain, displacement risks frequently emerge through identifiable patterns and warning signs, well before large-scale population movements occur. Many displacement-prone contexts also have decades of historical data on how conflict, fragility, socio-economic vulnerability and climate stress interact to drive population movement. These can be used to inform models for forecasting displacement, which do not necessarily seek to predict entirely new crises, but rather identify contexts where known drivers of displacement persist, intensify or interact in ways that have historically resulted in population movements.

Instead, the persistent gap lies in translating warning signals into pre-agreed actions and pre-arranged financing. Encouragingly, predictive capacities for conflict-induced displacement are advancing rapidly, while operational experience and community-level evidence increasingly demonstrate that anticipatory action in FCV settings is already taking place informally. Initiatives such as Danish Refugee Council’s (DRC) Anticipatory Humanitarian Action for Displacement (AHEAD) and Start Network’s work on anticipatory action for the impacts of conflict are working to translate signals into earlier action with emerging pathways for feasible scale-up. The question is no longer whether anticipatory action can happen, but what kinds of actions are feasible, appropriate and effective across different conflict dynamics and displacement patterns.

Communities are already anticipating – but formal systems are not

The DRC integrated context analysis for 2025, which spans eight conflict-affected countries, revealed a consistent pattern: communities already act on informal early warnings long before formal humanitarian systems respond. Households described relocating assets and family members, adjusting livelihoods, activating diaspora-support networks and moving pre-emptively based on locally understood signals that are often absent from humanitarian systems.

Yet, local knowledge alone does not guarantee positive outcomes. While communities may recognize warning signals and act upon them, they are not always in a position to choose the safest or most effective course of action. In the absence of resources, support or viable alternatives, households may resort to selling assets, accumulating debt, reduced food consumption or other negative coping strategies to manage rising risks.

Rather than replacing community foresight, anticipatory action can add value by expanding the range of options available before and during unavoidable displacement. Combining local knowledge with predictive analysis, displacement forecasts and pre-arranged support can help households take earlier, safer and more informed decisions. It can also broaden the reach of information and assistance beyond existing social, political or ethnic networks, helping to ensure that anticipatory action is guided by a shared understanding of risk, rather than access to information alone.

Not one crisis, not one model: designing for diverse conflict typologies

One significant recent shift is the move away from treating anticipatory action for conflict as a single operational model with a uniform level of feasibility and impact potential. Emerging practice under the DRC’s AHEAD initiative suggests that different conflict typologies generate different displacement dynamics, lead times and realities for humanitarian access; these determine which forms of anticipatory action are operationally possible.

• In settings with cyclical intercommunal violence, such as parts of Ethiopia and South Sudan, where escalation patterns can be anticipated locally and community mechanisms remain functional, prevention-oriented models are being tested. These draw on local peace infrastructure to reduce escalation risks before displacement occurs.
• In complex conflict environments, such as Myanmar, where active conflict and humanitarian-access constraints limit the possibility of preventing displacement, anticipatory approaches focus on community-led protection and safer movement through localized anticipation systems.
• Corridor models are being deployed along the South Sudan–Uganda displacement corridor to anticipate cross-border movement and support safer transit before humanitarian needs peak at destination sites.
• In contexts where conflict dynamics can escalate rapidly and humanitarian access is often constrained, such as Burkina Faso, Niger and Yemen, anticipatory approaches focus on identifying likely displacement movements and preparing for increases in arrivals before needs peak. The aim is to bridge the gap between early warnings and humanitarian response, enabling assistance to reach affected populations sooner.

The diversity of these approaches is what context sensitivity looks like in practice, with diverse FCV typologies requiring a corresponding variety of anticipatory modalities, trigger systems and feasibility framings.

A dual-trigger architecture for anticipatory action

An important operational development in anticipatory action for conflict is the use of more agile trigger methodologies. Rather than relying on single-path forecasting models, as used in climate-focused anticipatory action, these combine phased activations, context monitoring and scenario-building.

The DRC is scaling up a dual-trigger system that combines predictive analytics with community-based early warnings. Rather than relying on a single forecast or automated threshold, it bridges longer-range forecasting with real-time local intelligence. The first trigger (or layer) uses an AI-supported predictive model to identify districts projected to experience unusually large increases in displacement; for the second, communities identify signals that often precede violence and displacement, such as movement restrictions, market disruptions or increased activity by armed groups.

Start Network’s experience also highlights the importance of flexible ways to trigger anticipatory action in conflict settings. A recent evidence review found that effective early action often relies on qualitative analysis, scenario building and expert judgement, rather than quantitative triggers alone. In Lebanon, for example, agencies combined real-time field information and risk analysis with displacement scenarios to support decision-making under uncertainty. Its Conflict Anticipatory Action Project is now developing pilot frameworks in Somalia and South Sudan to test how scenario analysis, community foresight, predictive models, flexible triggers and pre-arranged finance can enable action before the humanitarian impacts of conflict are fully felt.

Debating anticipatory action for FCV impacts at the Nairobi dialogue

These themes surfaced during a strategic dialogue in Nairobi, Kenya, convened by the Dutch Ministry of Foreign Affairs and the DRC in May 2026. Among the donors, practitioners, regional institutions and refugee-led organizations present, there was strong agreement that conflict-induced displacement should no longer be treated as inherently unpredictable or outside the scope of anticipatory action. The challenge to changing this is institutional, rather than technical: whether humanitarian systems possess the financing flexibility, risk tolerance and agile decision-making structures required to act on the available signals before crises escalate.

Participants noted that while anticipatory action for climate shocks has, in part, increased in scale through standardized methodologies, triggers and activation mechanisms, the diversity and complexity of FCV contexts means that such standardization is neither feasible nor desirable. Instead, it will need to be achieved through adaptable frameworks that preserve contextual sensitivity, rather than through uniform approaches applied across fundamentally different crises.

The work that remains

Discussions have moved beyond whether anticipatory action for conflict-induced displacement is possible, towards practical discussions of what this looks like across different conflict dynamics and displacement trajectories. This shift comes at a critical time: forced displacement continues to rise, many FCV crises are increasingly protracted, and humanitarian funding is failing to keep pace with growing needs.

Against this backdrop, relying on costly post-displacement responses is increasingly difficult to sustain. In contrast, conflict-sensitive anticipatory action can reduce humanitarian impacts while preserving the agency, dignity and choice for affected populations – while also making more effective use of scarce resources. The challenge is not simply one of technical refinement, but about developing anticipatory approaches that are flexible enough to reflect the realities of conflict-affected contexts, grounded in community knowledge and capable of operating under uncertainty.