Climate Policy Radar
Climate Policy Radar is a free, open non-profit search platform that turns the world's climate laws, policies, litigation and climate finance documents into structured, full-text searchable evidence for policymakers, researchers, lawyers and analysts worldwide.
What is Climate Policy Radar?
Climate Policy Radar is a non-profit research platform that gathers the world's climate governance documents into a single searchable layer. Laws, national policies, court cases, climate finance records and submissions to the United Nations sit side by side in one corpus, with more than 35,000 documents in the public app and coverage spanning every party to the UN Framework Convention on Climate Change: 196 countries plus the European Union, along with Taiwan, Palestine and Western Sahara. Every document is the primary text itself, sourced from a named institution rather than summarised second hand.
The work happens in three stages. Documents are collected and monitored with partner institutions, then structured: subject experts agree which concepts are worth tracking, and small, purpose-built AI models classify those concepts across millions of pages, under human review at every step. Search then runs on that structure. A dense retrieval model works alongside fuzzy string matching, so a query for 'internal combustion engine' also surfaces 'ICEV', 'fossil fuel car' and 'gasoline car' without the exact wording. Documents published in other languages are machine translated into English, which is what makes a single query reach across the whole corpus.
Rather than one interface for everyone, the organisation runs seven. Climate Change Laws of the World, built with the Grantham Research Institute at LSE, covers legislation and NDCs. The Climate Litigation Database, with the Sabin Center for Climate Change Law, covers court cases. Climate Project Explorer tracks projects and money across the four largest multilateral climate funds, Rio Policy Radar covers submissions to the three Rio conventions, and the POWER Library, Global Stocktake Explorer and Climate Council Report Library each serve a narrower field.
Everything is free. There is no pricing page and no paid tier: search needs no account, and direct access to the full dataset through a data warehouse, an API and MCP support is advertised as free for non-profits. The methodology, the pipeline code, the trained models and the concept store behind every search are all published openly, which means the classification work can be audited rather than taken on trust.
What it does
- Search the full text of more than 35,000 climate governance documents from every country
- Compare how one concept is treated across jurisdictions, document types and years
- Run natural language queries that expand to related terms instead of exact keywords
- Explore national submissions to the three Rio conventions on climate, nature and land
- Look up climate-related court cases worldwide through the litigation database
- Follow projects and funding flows across the four largest multilateral climate funds
- Pull the data into your own systems through the API, MCP support or the data warehouse
When to use Climate Policy Radar / When not to
A quick filter to help you decide if Climate Policy Radar is the right fit.
When to use Climate Policy Radar
- Climate policy analysts and government negotiators comparing national commitments across jurisdictions
- Academic researchers and university teams working on climate law, governance, nature or land use
- Lawyers and litigators tracking climate-related court cases through the Climate Litigation Database
- Climate finance analysts following project pipelines and funding flows across the major multilateral funds
- Data scientists and NLP practitioners who want open climate models, datasets and pipeline code to build on
When not to use Climate Policy Radar
- Anyone after a general-purpose chatbot or writing assistant: nothing here generates text for you
- Legal researchers working outside climate, nature and land governance, where the corpus stops
- Companies wanting turnkey ESG reporting software rather than the underlying source documents
- Teams needing a non-English interface today, since search runs in English and translation points towards it
- Users expecting a real-time feed, as the database is curated and updated manually by expert teams
How to use Climate Policy Radar
A typical end-to-end flow, from setup to results.
- Open app.climatepolicyradar.org: no account is needed to search
- Type a question or a set of keywords into the full-text search box
- Borrow one of the suggested starting points, such as 'Latest NDCs' or 'Climate framework laws'
- Turn on the related phrases option so the search reaches beyond your exact wording
- Narrow the results by jurisdiction, region, year or document type
- Open a document to read the primary text, with its annexes, amendments and translations attached
- Switch to the specialised interface that fits your subject, such as climate-laws.org for legislation or the Climate Litigation Database for court cases
- Read the published methodology and codebook on GitHub before citing anything from the database
- For programmatic work, write to partners@climatepolicyradar.org about API, MCP or data warehouse access
- Report a missing or inaccurate document through the feedback form or support@climatepolicyradar.org
Pros & Cons
Pros
- Completely free, with no account required to search
- Primary sources from named institutions, not summaries or second-hand claims
- Human review claimed at every step of the AI classification work
- Methodology, pipeline code, models and training data all published and auditable
- Global coverage: every party to the UNFCCC, 196 countries plus the European Union
- The publisher documents its own model biases and limitations rather than hiding them
- Backed by reference institutions including LSE Grantham, the UNFCCC and the Sabin Center
Cons
- The corpus is updated manually, so recent texts can lag behind publication
- Interface and search are English only; a multilingual interface is described as future work
- No reuse licence is named on any accessible page, despite openness being the stated goal
- The terms and conditions page renders only in a browser, so its clauses are hard to check
- The semantic model is general-purpose and trained on short passages, so long queries fare worse
- No mobile app and no browser extension
- Data warehouse access is arranged by contacting the team rather than self-service
Pricing & Plans
Climate Policy Radar is free of charge and there is no lowest price point to report. The site carries no pricing page and publishes no paid plan: the tools and datasets are described as free and open, search in the public app requires no account, and direct access to the underlying data warehouse is advertised as free for non-profit organisations. The organisation is a non-profit funded by philanthropic grants rather than by subscriptions.
Data, GDPR & hosting
A consolidated view of how Climate Policy Radar handles your data.
GDPR overview
Climate Policy Radar publishes a dedicated privacy policy, last revised on 19 October 2022, which commits explicitly to the UK GDPR and the California Consumer Privacy Act. It sets out six data subject rights — access, rectification, erasure, restriction, objection and portability — and undertakes to respond within one month, through a single address, support@climatepolicyradar.org. A cookie banner allows analytics cookies to be rejected outright, and the company says it honours the browser Do Not Track signal even where its analytics vendors do not. It states that it does not knowingly collect personal data from children under 13. No Article 27 EU representative is named, which is consistent with a UK company operating after Brexit.
Who owns the data?
Climate Policy Radar CIC, a Community Interest Company limited by guarantee registered in England and Wales under company number 13377442, is the controller. Its privacy policy says personal data is collected only when you volunteer it, under explicit consent or another valid UK GDPR basis, and is shared only with stated categories of processors: cloud computing, analytics, storage, and CRM or mailing tools. Hotjar is contractually forbidden from selling anything collected on the company's behalf. The documents in the corpus remain the work of the named institutions that publish them. One caveat: the terms and conditions page renders only in JavaScript and could not be read, so any ownership clauses it carries remain unverified here.
Reuse rights
Openness is the stated purpose rather than a side benefit. Climate Policy Radar says its data, tools and systems are freely accessible and can be adopted, integrated and extended by others, and it backs that up by publishing its methodology, pipeline code, trained models, training data and concept store in public. Direct access to the full dataset runs through a data warehouse with API and MCP support, advertised as free for non-profits and arranged by writing to the partnerships team. There is a real gap, though: no reuse licence — Creative Commons, ODbL or anything else — is named on any page that could be read, and the terms and conditions did not render, so confirm the redistribution conditions directly before republishing the data or building a commercial product on it.
Data retention & training
Hosting summary
The privacy policy states that information is stored in secure, regulated data centres in the United Kingdom, Ireland and the United States. No hosting subcontractor is named: the policy describes categories of providers — cloud computing, data analytics, data storage, and CRM or mailing tools — rather than individual vendors. On the analytics side several third parties are named directly: Hotjar, Google Analytics, PostHog, YouTube and Retool. Hotjar processes a visitor's IP address during the session and stores it in de-identified form within a pseudonymised profile. One technical point worth separating out: the domain resolves to 31.43.161.6, an Amazon anycast node exiting in Amsterdam. That is a content delivery node, not a storage location, and it does not contradict the three jurisdictions the policy declares. The editorial site is served through Framer, while the search application runs separately on app.climatepolicyradar.org.
Things to keep in mind
Risks and trade-offs to weigh before adopting Climate Policy Radar.
- The corpus is updated by hand, so a blank result may mean the document has not been added yet rather than that no such law exists
- Documents are machine translated into English, and legal nuance can shift in translation
- The semantic model is a general-purpose one whose biases the team documents openly, so it can misread climate or policy specific concepts
- Search silently falls back to fuzzy matching on queries flagged as risky, which changes behaviour without any on-screen warning
- Treat the database as an entry point rather than an authority: check the official text before relying on a document legally
- No reuse licence is named, so republishing the data without confirming the terms carries a real risk
- The service depends on philanthropic funding, so its continuity is not contractually guaranteed
Setup & Integrations
Technical difficulty
For ordinary use there is nothing to set up: open the app, type a query, and read the results. No account, no installation, no configuration. The real effort sits elsewhere — in reading the published methodology and codebook closely enough to cite the data correctly, which matters for anyone quoting it in research or legal work. Programmatic use is a different matter: the API, MCP support and data warehouse require a prior conversation with the team and genuine technical skill. Training sessions are offered to government officials and researchers, online and in person.
Deployment
Supported languages
Behind Climate Policy Radar
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
Does Climate Policy Radar cost anything?
Do I need an account to search?
What is actually in the database?
Which countries are covered?
What languages does it work in?
Is there an API?
Where do the documents come from?
Does the AI make decisions on its own?
Who runs Climate Policy Radar?
Where is my data stored?
Should you pick Climate Policy Radar?
Climate Policy Radar behaves less like a product than like public infrastructure, and that is the right way to judge it. A British Community Interest Company founded in 2021, funded by philanthropic foundations rather than subscriptions, it has put more than 35,000 climate governance documents behind one search box and made the whole apparatus inspectable: methodology, pipeline code, trained models and the concept store are all public. Its data partners are the institutions that hold the source material in the first place, from the UNFCCC and the CBD to the LSE Grantham Research Institute and the Sabin Center, and it reports more than a million users a year.
What makes it credible is also what limits it. The corpus is curated and updated by hand, which the organisation states plainly, so an absence in the database is not proof of an absence in the world — a distinction that matters when the question is whether a country has legislated on something. The semantic model is a general-purpose one trained on short passages, and its biases are documented by the team itself rather than discovered by users. The interface and search work in English, with source documents translated towards it; a multilingual interface is described as future work.
Two gaps deserve attention before serious reliance. No reuse licence is named anywhere on the accessible pages, despite openness being the stated purpose, and the terms and conditions page renders only in a browser, so its clauses could not be checked. The privacy policy has not been revised since October 2022. None of this undermines the research value, which is considerable and hard to find elsewhere, but anyone planning to republish the data or build a product on it should confirm the terms directly first.
- Choosing a selection results in a full page refresh.
- Opens in a new window.