Back to Newsroom

RiskThinking.ai Launches Open-Source Climate Risk Engine Free for Noncommercial Use Worldwide

The new Climate Risk Commons platform provides a shared, open-standard computing framework to democratize and increase transparency in physical climate risk measurement worldwide.

News provided by
RiskThinking.ai
November 17, 2026
11:30 am
EDT
Dr. Ron S. Dembo, founder and CEO of RiskThinking.ai and founder and chair of the Riskthinking Foundation / Source: RiskThinking.ai (EZ Newswire)

TORONTO, Canada, November 17, 2026 (EZ Newswire) -- RiskThinking.ai has announced the launch of Climate Risk Commons (CRC), an open-source framework that facilitates the democratization of physical climate risk and redistributes analytical tools to the wider research community.

The full launch has been preceded by beta releases in September that also saw the creation of a public data specification, an open-source Python SDK, a reference computation pipeline, and a canonicalized hazard dataset from RiskThinking.ai. Data can be compared across platforms with a common schema for public climate hazard data, academic models, AI weather models, and commercial vendors, all with the same risk analytics.

Data incompatibility represents a major bottleneck in climate-related finance: providers of climate hazard data release their output in various formats, grids, and measurement units, which makes the data unusable and not compatible with financial risk models.

Developing a Common Standard for Climate Risk Data

Climate Risk Commons isn’t meant to be another proprietary dataset. An open standard will support data providers in collaboratively sharing, structuring, and using physical climate hazard data such as flood and drought data. It is a complete analytic engine, the same one used by RiskThinking.ai's largest clients, that generates risk measures from the data.

The new release contains the CRC Framework, a Python Software Development Kit (SDK), data adapters, starter impact functions, and reproducible notebooks. Data adapters are available for ERA5, JRC, CMIP6, and RiskThinking.ai. In addition, open datasets are added weekly as users create adapters for the datasets they wish to use or contribute.

The tool lets users load hazard data, apply damage functions, and rate a single asset or portfolio against a return-period hazard level. Researchers can load their own valid data, compare models, and develop new damage functions while preserving the full fidelity of the analytical pipeline across any data source.

Details about the infrastructure are published on the project's website, climateriskcommons.org.

Making Climate Risk Assessments Comparable

Banks, insurers, asset managers, and regulators face risk when climate risk providers produce different results for the same property or portfolio. Variations in formats, hazard models, and methods can create confusion around key drivers.

Climate Risk Commons offers a common standard, the first of its kind, to compare sources side by side. With reproducible notebooks, users can compare various hazard sources to the same assets and see where they agree or differ.

It also permits the inclusion of multiple hazards for the same asset and calculates value at risk (VaR) and conditional value at risk (CVaR). The output report explains how the results were calculated so that a user can check data, assumptions, and calculations.

Extending Access Through Open-Source Technology

Dr. Ron S. Dembo, founder and CEO of RiskThinking.ai and founder and chair of the Riskthinking Foundation, said the initiative is intended to extend access to climate risk analytics beyond institutions that can already afford sophisticated tools.

“If this data and analytics only serve the institutions that can already afford it, we haven’t solved the climate-risk problem. We have just priced some people out of knowing it,” Dembo said.

The architecture, which is built with an AI-first approach, offers agentic capabilities in the form of workflows. This allows AI agents to execute climate risk assessments on prompt-based natural language without having to write any code or master the data format prior to running an assessment.

The project offers code and documentation on GitHub, a Discord community for users, and webinars and other events that allow researchers, engineers, and risk teams to work together.

Open Software with Separate Licensing Terms

The Climate Risk Commons software is released under AGPL-3.0-or-later, permitting both commercial and noncommercial use under the license terms. The CRC Open dataset is licensed under CC BY-NC 4.0, while third-party datasets retain the licensing terms under which they were provided.

RiskThinking.ai will continue to offer commercial data and analytics alongside the open framework. These offerings include a database covering more than 7 million assets, over 2,000 pathway scenarios, a patented multi-hazard correlation engine, and enterprise services. These offerings remain subject to separate commercial agreements.

With the full launch of Climate Risk Commons, RiskThinking.ai aims to establish a shared technical foundation for measuring physical climate risk, enabling researchers and institutions worldwide to compare climate hazard data and develop more consistent risk assessments.

About RiskThinking.ai

RiskThinking.ai is a climate risk analytics company specializing in measuring physical climate risk. The company provides climate risk data, modelling, and analytical tools for organizations assessing the climate-change exposure of assets or portfolios.

Media Contact

Ron S. Dembo
Founder and CEO, RiskThinking.ai, Riskthinking Foundation
rdembo@riskthinking.ai