The 2026 Guidebook: Is disaster risk worth ignorance?
Sundeep Reddy Mallu, Co-Founder CTO Resilience AI
The Maps Said It Would Not Rain This Much
CoreWeave is a major AI Cloud provider. Persistent heavy rain and strong winds at their site in Denton, Texas. This “low-tech” environmental nuisance prevented contractors from pouring concrete foundations required for a massive data-center complex. The 60-day delay on-site caused ripple effects on the 260-megawatt project. This delay caused the overall project timeline to be pushed out by several months.
Two months of construction delay caused by heavy rainfall led to nine months of delay in completing the data center. Could this be prevented?
The consequence is a direct loss and a cascading impact. The data center commitment of 10 GW new capacity in the USA to handle massive projected AI workloads was delayed. Projected to cost $800 billion to build was at risk. CoreWeave lost a staggering $33 billion in market valuation in 6 weeks (46% drop in stock price). Firms tied to datacenter; the likes of Oracle and Broadcom saw their shares drop.
In 2024, cascading impact of unmanaged disaster risk by companies and government was USD 2.3 Trillion.
Globally the occurrence, the ferocity and impact of naturally triggered disaster events, heatwave, cold wave, landslide, earthquake, floods, wildfire, tornado, tropical cyclones is omnipresent in 2025.
- Per WMO, 2025 set to be third warmest year on record globally
- Flooding is projected to remain top weather-related supply chain risk for 2025
India recorded extreme weather events on 331 of 334 days till November 2025
This is a new reality. Businesses are deeply physical systems – dependent on built-environment, land, power, water, data network and execution schedules that are increasingly exposed to extreme event volatility. Climate risk limited to climatic parameters of rain, heat, wind, rising sea level are no longer adequate as “edge risks”. Naturally triggered disasters are beyond climate risk as “operational risks”, that warrant greater scientific attention intersecting with systemic planning.
At Resilience AI, with an ear to the ground we have seen this movie playout beyond data center to nearly every sector and a very high underreporting of business losses when a disaster disrupted financial muscle, physical assets, market value.
Two additional sectors are rail network and insurance, familiar to most practitioners, yet an unsolved riddle. Tropical Cyclone Sean led to flooding of rail infrastructure in Western Australia’s Pilbara region. The region recorded 274.4 mm of rain in a 24-hour period on January 20,2025. Rio Tinto’s iron ore railcar dumper facility at East Intercourse Island shutdown for 4 weeks with a 17% quarter-on-quarter drop in shipments. Southern California wildfires in January 2025 burned down more than 55,000 acres and damaged/ destroyed thousands of structures in its wake. This single fire contributed to massive insured losses, amounting to $40 billion. Facing concentrated claims from one single region affects combined ratios, stress on company reserve and future premium increases.
This trend is repeated in earning calls of companies in S&P 500 Index
- CSX (Transportation/Rail) – Weather-related disruption costs adding up to $120 to $125 million
- Walt Disney Company (Media) – Hurricanes Milton and Helene had about $120 million impact on experience segments at their parks in Florida
- Xcel Energy (Utility) – Marshal wildfire settlement led to $290 million charge in 2025
- HCA Healthcare – $250 million hit from hurricanes
The real question in 2026, could this be tackled better?
What should companies and governments invest in for business and service continuity – siloed tools such as climate intelligence, weather alerts, crisis and/ event management or a systemic disaster management system?
Businesses are not factoring “disaster risk” in performance metrics and business planning such as business continuity plans (BCP), annual operating plan (AOP), strategic plans (STRAP). Governments are myopic with project-based funding, 80% investment towards response and recovery and 20% towards preparedness. Small and medium businesses do not even consider extreme event planning until livelihood is loss.
The real difference between sustainability report and a disaster management plan in a business and similarly, between disaster response fund and disaster management plan in a government is the treatment of disaster risk.
Can disaster risk be an inherent macroeconomic indicator across business functions for 365 days preparedness similar to using GDP, Inflation rate, CPI. Can a scientifically derived “Resilience Factor” be part of the Risk Index equation such as FEMA NRI (National Risk Index) and India DRI (Disaster Risk Index), towards informed development planning and disaster budget allocation.
All these incidents reinforce our view at Resilience AI, that naturally triggered disaster events can cascade into billion-dollar consequences for businesses. In the age of AI and enterprise software, resilience is not just about uptime (in a business function) and emergency and/or crisis response (in a government function). It is about building the underlying infrastructure that can withstand an increasingly unpredictable planet.
The Texas flood tells a chilling story. Overnight 3 into 4 July, a sudden deluge turned Texas Hill Country into a disaster zone, killing more than 100 people and leaving dozens missing. In a few hours, 10-18 inches of rain swamped the Guadalupe River basin, sending the river surging 26 feet in just 45 minutes. Many of the victims were young girls at a summer camp, caught unaware as floodwaters tore through sleeping quarters around 4 AM. Although the US National Weather Service issued warnings ahead of time, local sirens were lacking and the final alerts came when most were asleep.
Flash floods killed 706 people and injured 965 others in Pakistan since 26 June. At least 344 fatalities occurred from 14-16 August, including 328 in Khyber Pakhtunkhwa. The floods of 2025, caused by intense monsoon rains and the release of water from dams in India, have submerged Punjab, Khyber Pakhtunkhwa, and other regions. Over 2,200 villages are now underwater, and 750,000 people have been evacuated.
In May and June 2025 alone, three glacial outburst floods hit Nepal, Afghanistan and Pakistan, with two more in Nepal on 7 July. If warming continues, the risk of such floods could triple by the century’s end.
Vietnam faces its own battles. While specific 2025 flood data was limited in my search, northern India, Bangladesh, Bhutan, Nepal, Thailand, Laos, and Vietnam experienced heavy to very heavy rainfall, with localized areas exposed to extreme downpours.