Mike D’Aurizio and Christophe Defert

In January 2025, the devastating wildfires that tore through Los Angeles destroyed almost 12,000 homes and killed 31 people [1] (while potentially contributing to another 400 deaths [2]). Over 175,000 people were forced to evacuate their homes, and 70% of those displaced had still not returned a full year later [3]. This was the costliest wildfire disaster in US history, with total economic losses estimated to be at least $53bn [4].

Unfortunately, this event was not just a random outlier. It was more the starkest illustration yet of a growing trend: natural catastrophes resulting from rising global temperatures.

The past 11 years have been the 11 warmest in the 176-year observational record. For the first time ever, the three-year average temperature for 2023-2025 was 1.5°C above pre-industrial levels [5] – the supposed limit enshrined in the Paris Agreement. In 2025, natural catastrophes caused $224bn in economic losses, while insured losses exceeded $100bn for the sixth consecutive year [6]. The effects of climate change are no longer hypothetical; they are already happening, with tangible social, environmental and economic costs.

Much of the debate around climate has focused on mitigation: reducing emissions to prevent further warming. This is essential work (and continues to account for most of our own investment activity). But at the same time, we also need to address the consequences of our failure to act faster: the global warming already baked into the climate system, which will intensify for decades.

Investing in adaptation – building the capacity to withstand, respond to, and recover from climate impacts – is not just a necessity in its own right. It’s also a precondition for the transition more broadly, because communities and overwhelmed by climate disruption will struggle to transform in the way decarbonisation requires. Thankfully, the rapid advances in artificial intelligence and data science are now giving us new tools to improve adaptation and resilience.

The intelligence opportunity

Adaptation is, at least in part, an information and risk management problem.

For most of history, climate risk assessment has relied on imperfect historical data: flood maps drawn from past events, actuarial tables calibrated to previous loss patterns. That approach is increasingly inadequate in a world where the climate is moving beyond the realm of past human experience.

AI and data science are enabling a more forward-looking, granular, and dynamic approach to risk management. This can potentially unlock a substantial economic opportunity: BCG estimates the total market for AI-powered climate risk modelling could be worth $75bn per annum by 2030 [7].

Insurance is one sector where the case for investing in adaptation is compelling. Every dollar invested in adaptation today potentially generates more than ten dollars in benefits over the next decade [8]. Of the $417bn in natural disaster losses in 2024, only $154bn was insured – leaving 63% unprotected [9]. And critically, uninsured losses fall disproportionately on the communities with the fewest resources to absorb them.

Delos Insurance, a US-based company we backed at Series A in 2024, shows how better climate intelligence can create scalable market solutions. Delos uses AI and remote sensing data, combined with proprietary wildfire science, to score catastrophic wildfire risk at the individual property level – a granularity that traditional actuarial models simply cannot achieve. They can also update their wildfire models dynamically to respond to changing physical risks. This AI-driven approach to measuring wildfire risk allows Delos to offer home insurance to households in wildfire-exposed markets that incumbent insurers have abandoned, while also providing policyholders with personalised guidance on reducing their specific risks. The impact is tangible: millions of California families who had lost access to affordable insurance can now be served affordably, with intelligently priced coverage that can be the difference between financial resilience and ruin.

The same logic applies across a range of adaptation applications. AI-powered early warning systems are extending the accuracy and lead time of extreme weather predictions, enabling pre-emptive action rather than reactive response. In agriculture, for instance, precision forecasting helps farmers make better planting, irrigation, and harvesting decisions, reducing both yield losses and water use. Applied to infrastructure, predictive intelligence allows operators to reroute power flows ahead of a storm, position maintenance crews before equipment fails, and manage water resources before a drought becomes critical.

What unites these applications is a shift from managing climate risk retrospectively – through disaster relief and recovery – to managing it proactively, through better data and earlier decisions. The economic rationale for this is clear and compelling.

Where intelligence meets impact

For investors like us, what’s particularly attractive about adaptation and resilience as an investment theme is the clear convergence between financial return and social value. The businesses best placed to succeed are those that help customers – whether that’s uninsured homeowners, farmers, infrastructure operators, or city planners – to survive and thrive in a more volatile climate.

There are also positive regulatory tailwinds. Physical climate risk disclosure requirements are tightening across both the EU and the US, creating demand for the data and analytical tools that make credible disclosure possible. Companies that cannot quantify their climate exposure will increasingly face both regulatory and investor pressure to do so. But it’s still early days: the tools to manage climate risk are still being built, the datasets required to power them are still being assembled.

For impact-driven investors, this combination – a large and growing addressable market, clear commercial logic, an early-stage competitive landscape, and measurably better outcomes for underserved populations – is compelling. We must not falter in our attempts to reduce current and future emissions. But thanks to the intelligence era, we also have an opportunity to improve our adaptation and resilience to climate change that is already happening. Getting this right will be an essential part of accelerating the transition to a more sustainable future.

[1] https://www.nature.com/articles/s43017-026-00793-z

[2] https://www.bu.edu/sph/news/articles/2025/death-count-for-2025-la-county-wildfires-likely-hundreds-higher-than-official-records-show/

[3] https://www.calfund.org/news-and-events/a-year-after-la-wildfires-most-survivors-are-still-displaced-and-struggling-financially-department-of-angels-survey-finds/

[4] https://www.munichre.com/rmp/en/the-re-brief/risk-adaptation/what-if-the-next-california-scale-wildfire-happens-in-the-midwest.html

[5] https://wmo.int/news/media-centre/wmo-confirms-2025-was-one-of-warmest-years-record

[6] https://www.munichre.com/en/company/media-relations/media-information-and-corporate-news/media-information/2026/natural-disaster-figures-2025.html

[7] https://www.bcg.com/publications/2026/the-capital-opportunity-in-ai-enabled-sustainability

[8] https://www.wri.org/news/release-wri-study-finds-climate-adaptation-investments-yield-massive-returns

[9] https://www.ajg.com/gallagherre/-/media/files/gallagher/gallagherre/news-and-insights/2025/natural-catastrophe-and-climate-report-2025.pdf