Rethinking humanitarian response in the age of compounding crises and better intelligence

The humanitarian system faces a brutal math problem: needs, crises, demands are growing faster than funding. Is there a smarter way forward?

Farmers abandoning fields in Zambia as pre-planting forecasts turned into failed harvests. Dadaab’s outskirts filling with makeshift shelters as Somalis escape the prolonged drought at home. County roads in northern Kenya washed out or parched, as communities alternate between flood and drought in a single year.

Each of these climate-driven emergencies triggers the same pattern: we mobilize heroic responses with emergency airlifts, surge capacity, and round-the-clock logistics. These are very heroics we rightly celebrate every year on World Humanitarian Day. But heroics after disaster will never be enough. And while our response spending grows, human need grows faster, and donor funding is slimming down.

The UN’s Global Humanitarian Overview shows we need more than US$45 billion to reach 181 million people in need in 72 countries. Despite growing needs, humanitarian funding plunged in 2025, falling 39% below 2024 levels, with major donors contributing significantly less.  World Humanitarian Day has always been about solidarity with those who suffer. With today’s advancements in science and intelligence, solidarity would mean smarter interventions before disaster strikes, not after.

The Mathematics of Prevention

Anticipatory action uses intelligence to act before disaster strikes. When clear warning signs appear, pre-approved funds are released automatically and essential supplies are pre-positioned based on forecasting models.

In today’s climate of slim funding, rather than parallel warning systems, we need layered triggers where the same drought forecast activates livelihood support from development actors at the first threshold, then humanitarian surge capacity when conflict indicators cross secondary thresholds. This is how we do more with less.

FAO research shows 7:1 benefit-cost ratios for acting ahead of droughts, yet anticipatory action represents less than 1% of international humanitarian assistance. Under shrinking budgets, this is mathematically irrational.

Real-world examples prove the model works.

Consider the 2024-25 El Niño season in Southern Africa – Malawi received a $3.38 million ARC sovereign payout; modest but well timed. Instead of waiting for hunger headlines, cash reached households during the lean season while markets were stable.

In Kenya, the Red Cross operationalized Early Action Protocols with impact-based forecasting, shifting resources to at-risk counties as trigger thresholds activated. Across the border in Uganda, households received early cash for tools, sandbags, and community warnings which helped divert water and speed up recovery.

Perhaps most promising: conflict early-warning systems are getting smarter about the climate-security nexus. A joint study found that healthier vegetation and higher rainfall are associated with lower odds of physical conflict the following month. Such progress aims to turn forecasts into preventive measures (via targeted dialogues, safe-passage coordination, protection deployments) before dry-season pressures explode.

So why is this sort of approach not yet mainstream? Mainly because we’re still building parallel systems instead of integrated ones. Climate services, humanitarians, and peacebuilders still work in parallel, when stronger bridges between meteorological services, cash working groups, and protection clusters are both practical and overdue. Second, metrics still count outputs instead of avoided losses. We should be publishing avoided-loss figures and cost–benefit analyses, following WFP/Anticipation Hub guidance, not just post-distribution monitoring.

The good news is practitioners have a growing palette of functioning toolboxes: regional seasonal outlooks from ICPAC, food security early warning from FEWS NET, and market monitoring to calibrate cash transfers. Platforms like our CGIAR Climate Security Observatory overlay climate, land, and water data with fragility indicators to help governments and agencies target anticipatory actions, with deep country-level analytics and partner interfaces. The insights from the CSO have helped partners point budget lines and safeguards at the right places sooner.

When climate forecasts, market dynamics, and conflict trends feed into one integrated risk picture, humanitarian teams and development planners can work from the same intelligence, preventing the costly duplication of parallel assessment systems.

The scale of the humanitarian crises today demands a shift in how we prepare and respond even as we enter the age of enhanced intelligence. These tools and services must reach the hands of county officers, national and regional decision makers, and multilateral agencies so they can act with adequate intelligence. At the Climate Security team of the Alliance of Bioversity International and CIAT, we support this shift with tools and analytics that so far have informed Kenya’s National Climate Change Action Plan, UNHCR’s Strategic Plan for Climate Action, and have guided strategies at IGAD and the African Union.

A Call to Advance Humanitarian Response

At COP30 in Belém this November, as negotiators will finalize the new collective quantified goal (NCQG) on climate finance we must carve out space for forecast-based action in fragile settings through sovereign risk pools, contingency financing, and adaptive social protection tied to objective, locally relevant triggers.

Reallocation needs to move up the agenda. Humanitarian needs are outpacing available resources so if there was ever a moment to reallocate, it is now. Science is providing the evidence and the mechanisms exists. What remains is the will to act and speed up early action to avert or mitigate disasters.

Authors: Peter Laderach and Grazia Pacillo, Co-Leads, CGIAR Climate Security, Alliance of Bioversity International and CIAT. Photo Credit: Georgina Smith / CIAT