EU Framework for Climate Resilience: What the Proposed Law Asks of Risk, Planning, and Investment

On 27 October 2026 the European Commission presented an EU Framework for Climate Resilience. The package has two core documents: a Communication, COM(2026) 720 (A Resilient Europe for a Changing Climate), and a proposed regulation, COM(2026) 730 (the European Climate Resilience Law, procedure 2026/0336 (COD)).

The law is a proposal. It would apply directly in every Member State if Parliament and Council adopt it, but the text can still change. The Commission has said it will start applying the approach in its own work before adoption. Reading the package as already-binding corporate law would be a mistake. Ignoring it until 2030 would also be a mistake, because the duties land first on public authorities whose permits, budgets, and tenders shape private projects.

This note is about what the proposal actually contains, how it reaches companies without a new corporate reporting obligation, and where a single climate-risk score is a poor fit for what the text asks assessors to show.

What the Commission is trying to fix

The proposal starts from a diagnosis the Commission has repeated since its 2023 assessment of national adaptation progress: strategies exist, implementation does not. Member States use different warming assumptions, different cycles, and different sector coverage. Shared risks — rivers, grids, supply chains, health — are then managed as if they stopped at a border.

The Communication is explicit about the shift it wants: from planning on past climate to planning on projected climate; from diffuse responsibility to named risk owners; from reacting after damage to managing risks that can already be foreseen; and from treating adaptation as a short-term cost to treating it as spending that avoids larger losses later.

Europe, the Commission says, is warming at about twice the global average. It cites economic losses from climate extremes of about €50 billion a year over 2022–2025 (about €200 billion in four years), against €8.6 billion a year in the 1980s. It estimates adaptation investment needs at about €70 billion a year, and says those investments can often return two to ten times their value in avoided damage. It also cites a global market for adaptation solutions projected at about €3.5 trillion by 2050. Those figures are the Commission’s, drawn from the studies named in COM(2026) 720. They are not a forecast of what any one organisation will save.

The same Communication is frank about insurance. Historically, only about a quarter of natural-catastrophe losses in the EU have been insured, on EIOPA and ECB figures the Commission cites. A framework that clarifies risk and still leaves most losses uninsured has not finished the job. The World Resources Institute made the same point in its statement on the package: direction is not the same as finance and accountability.

The legislative core

COM(2026) 730 would amend the Governance Regulation, (EU) 2018/1999, and operationalise the adaptation provisions in Article 5 of the European Climate Law, (EU) 2021/1119. It does not replace sectoral rules on water, buildings, critical entities, or transport. Recital 8 says it does not alter those obligations.

Five design choices matter more than the surrounding programme announcements.

1. A reference warming trajectory, not a new mitigation target

Article 4 and Annex I set a reference global warming trajectory for risk assessment, planning, and decisions:

Horizon	Reference warming above pre-industrial (1850–1900)
2030	1.5°C
2050	2°C
End of the 21st century	3°C

This is a planning baseline, not a statement that the EU has given up the European Climate Law or the Paris Agreement. Recital 9 says the trajectory does not affect the binding climate-neutrality objective, and that it should be used together with stress tests against more adverse scenarios.

Article 5 then requires assessments to consider alternative futures, extreme scenarios, and low-probability, high-impact risks, including tipping points. Article 12 lets the Commission update the trajectory by delegated act when the science changes, taking advice from the European Scientific Advisory Board on Climate Change.

France already plans against a national warming path that is hotter than the global average, because Europe warms faster. A shared EU baseline does not remove that national judgement. It stops neighbouring administrations from quietly using incompatible futures for the same river or the same grid.

2. Named policy risk owners

Article 7(6) requires each Member State to identify, for every major climate risk, one or more public authorities responsible for managing it. Annex IV asks the national strategic plan to name those authorities and to describe how climate resilience is built into public functions at national, regional, and local level.

The point is institutional, not rhetorical. The Commission’s own impact assessment history, restated in the proposal, is that adaptation plans failed where nobody had the mandate, the budget, or the technical capacity. “Risk owner” here is a public authority, not a chief risk officer inside a company. Companies will still be asked, by those authorities, who on their side can answer for a site, a concession, or a supply contract.

3. Assessments and one plan, on a fixed calendar

The proposal replaces four current planning instruments — national adaptation strategies and plans under the European Climate Law, plus the adaptation parts of national energy and climate plans and long-term strategies — with one national strategic plan. Reporting of progress stays inside the Governance Regulation, which the Commission presents as a simplification rather than a new reporting stack.

Deliverable	Who	First deadline in the proposal	Then
European climate risk assessment	European Environment Agency	30 June 2028	Every 5 years
Union strategic plan	European Commission	30 September 2028	Update by 30 September 2033, then every 5 years
National climate risk assessment, including subnational level where applicable	Each Member State	30 June 2029	Every 5 years
Integrated reporting on national climate resilience policies	Each Member State, under the amended Governance Regulation	15 March 2029	Every 2 years
National strategic plan on climate resilience	Each Member State	31 December 2030	Every 5 years
Statistics on deaths and economic losses attributable to climate change	Eurostat, from Member State data	31 December 2030	Annual
Commission review of the regulation	European Commission	30 September 2033	—

From 2028 the Commission also intends to assess Member State progress every two years through the European Semester. The five-yearly EU-level assessment under the European Climate Law would remain, including for Paris Agreement stocktakes.

National plans must be prepared with early public participation and with subnational authorities. Article 8 requires consultation of Member States likely to be significantly affected, at least neighbours, on transboundary risks and cooperation. A strategic environmental assessment under Directive 2001/42/EC can satisfy that duty if it actually covers the same ground.

4. Thirteen policy areas, and a statement of residual risk

Annex III lists the policy areas that both assessments and strategic plans must address:

Spatial planning
Critical energy systems
Critical transport systems
Critical communication systems
Water security and critical water infrastructure
Buildings
Public health
Food security
Economic and social policies
Budgetary policies
Financial systems
Crisis preparedness, response, and solidarity mechanisms
Ecosystems, ecosystem services, and biodiversity

Annex II tells assessors to identify major risks across near, medium, and long term; to judge whether current and planned policy can manage each risk; and to set out options, including how they perform under the reference trajectory and under more adverse projections.

Annex IV then asks plans to state the tolerable level of residual risk: the climate-related loss expected to remain after the measures, accepted by the decision-maker at the time, given uncertainty and financial, technical, and political constraints. Measures must name a responsible authority, a timeframe, expected costs and benefits (including non-monetary ones, where proportionate), and how maladaptation is avoided. “Maladaptation” is defined in the proposal: action that raises risk, shifts vulnerability, increases emissions, or leaves people worse off, now or later.

That is a different question from “what is our climate score”. It is “which losses are we prepared to accept, on what evidence, and which measures fail if the future is worse than the baseline”.

5. Just resilience, without binding quantitative adaptation targets

Plans must consider people in vulnerable situations and regions hit harder than others. The Communication says the Commission intends to bring that into the European Semester and into the strategic plans, including a common method for distributional impacts and for preparedness in social protection.

What the proposal does not do is set legally binding, quantified adaptation targets for 2030 or 2040. The Commission has framed limited legal obligations as consistent with its better-regulation agenda: common outcome indicators rather than a new target architecture. Annex V starts with four Union-level indicators:

deaths attributable to climate change
economic losses attributable to climate change, in euro, disaggregated where feasible
the share of those economic losses that is uninsured
climate-related contingent liabilities and fiscal costs

Civil-society groups, including CAN Europe and the European Environmental Bureau, had asked for interim targets, stronger funding earmarks, and a harder line against harmful subsidies and maladaptation. E3G described the package as laying foundations rather than finishing the system. That argument will be part of the legislative debate. It is also a live question in our work on the EU Better Regulation consultation: whether simpler process still shows if a policy will achieve its outcome. Four loss indicators can make damages comparable. They do not, by themselves, say whether a national plan reduced risk for the people most exposed.

Where companies meet the proposal

Law-firm commentary, including Eversheds Sutherland, has been careful on this point. The operative duties sit on the Union and the Member States. Articles 4(2) and 5(5) say Member States shall encourage public authorities to take the trajectory and the risk assessments into account when they exercise public functions, in particular budgeting, spatial planning, and environmental assessment under the SEA Directive (2001/42/EC) and the EIA Directive (2011/92/EU). “Encourage” is softer than a strict duty on every permit writer. It is still the channel through which a warehouse, a port concession, a housing scheme, or a grid upgrade will be asked for climate evidence.

Four paths are already visible in the package.

Public procurement. COM(2026) 720 points to the Commission’s September 2026 proposal to revise public procurement law (COM(2026) 590), which would treat climate resilience as a quality criterion. The Communication notes that public procurement is about 14% of EU GDP. That proposal has its own legislative path. Bid teams in construction, energy, water, transport, and facilities should assume buyers will start asking for heat, flood, and drought performance before the resilience law itself is adopted.

Permits and plans. Infrastructure, real estate, and energy projects that already need an environmental impact assessment or sit inside a spatial plan are the first place the reference trajectory becomes a document in a file. The proposal does not invent a new climate-permit regime. It tells authorities to use one warming path, and the national risk assessment, when they do work they already do.

Sector rules that already exist. The Communication spends a long chapter embedding resilience in instruments that are not waiting for this regulation: the Critical Entities Resilience Directive, TEN-E and TEN-T climate proofing, the Energy Performance of Buildings Directive’s overheating provisions, water and flood plans, the Nature Restoration Regulation, ports, health, and occupational heat. A company already inside one of those regimes should read the framework as a coordinating layer, not as a replacement.

Finance, valuation, and insurance. The Commission says revised EU Taxonomy delegated acts should make resilience investments easier to identify. The EIB Group plans at least €30 billion for climate adaptation from 2026 to 2030, and tools including a forthcoming open Climate Hazard Screening and Exploring Tool (ClimateSET). The Communication also says physical climate risk should show up more clearly in valuations and accounts, and that the Commission will keep supporting companies on transition and resilience plans. Separately, a Climate Insurance Alliance is meant to deliver a first set of measures by the end of 2027 on uptake, prevention, and capacity. EIOPA is to develop an Observatory for Catastrophe Risk Insurability and extend its PROTECT tool. None of that is a new CSRD datapoint. It is a change in what lenders, insurers, and public banks will have a common language for.

If the next decision is operational rather than legislative, the practical counterpart to this note is EU Climate Resilience Framework readiness: what to screen, what evidence a permit or tender file needs, and where a public hazard map stops being enough.

What a conforming assessment cannot honestly be

Annex II is uncomfortable for any product that collapses a site into one number.

A national or project assessment that “fulfils” the proposal has to show hazards, exposure, and vulnerability; systemic risks, including transboundary, cascading, and compounding risks; people and places that are hit harder; more than one future; and the residual risk left by each option. Finer maps do not satisfy that list. A confident loss figure to four decimal places does not either.

Two findings we have already written up sit directly on this text.

Vendor studies do not show a stable asset-level answer. The Investor Leadership Network’s comparison of seven physical-risk vendors found agreement on flood exposure near Paris that was statistically indistinguishable from chance, and no consensus on the top hazards for the test properties. GARP’s 13-vendor benchmark found the same pattern, including a geocoding error of 1,507 kilometres. We summarised both in Why Climate Risk Vendors Disagree. A score that cannot survive a second vendor is a weak annex to an environmental impact assessment.

Deep uncertainty is the condition Annex II describes, whether or not the regulation uses that phrase. When decision-makers do not share a probability over futures — emissions policy, tipping points, what counts as an acceptable loss — a single expected-value score answers the wrong question. Decision Making under Deep Uncertainty asks which options hold across many futures and which observable sign should trigger a change of plan. We set out the method, and its limits, in Beyond the Single Score. That remains a research topic for us, not a product we claim to deliver off the shelf. The regulatory text now asks public plans to do a version of the same thing: stress-test options, and say what residual risk is tolerated.

Resolution is part of the same honesty. The new European Climate Viewer turns the reference trajectory into local hazard information from Copernicus climate services. That is the right public baseline. It is not, by itself, a flood depth at a loading dock or a heat stress metric for a ward. What the Viewer shows, and where it stops. Which downscaling product is used can move projected economic damages by several percent of regional output, sometimes by more than the choice of climate model. We wrote the practical implications in Preparing Your Own Climate Data. The tools launched with the framework, and where they stop, are covered in EU climate resilience tools and enterprise data.

There is a distributional version of this gap, which we argued in Climate Risk Data Quality Is a Market Problem. If only well-funded buyers can purchase fine-grained risk data, and that data is not reliably better, then “resilience by design” will be precise for some balance sheets and coarse for the municipalities Annex III also covers. The proposal’s public tools and the EEA’s role as a centre of excellence are a partial answer. They do not close asset-level method questions.

Finance without a new pot inside the regulation

The legislative financial statement attached to COM(2026) 730 does not create a dedicated adaptation fund. Staffing for the Commission, the EEA, and EIOPA is described as redeployment, with figures for the next EU budget left indicative. The Communication points instead to instruments that already exist or are in parallel negotiations: the 2028–2034 multiannual financial framework, national and regional partnership plans, Horizon Europe, LIFE, the European Competitiveness Fund, the Social Climate Fund, State aid rules, and EIB programmes.

Two innovation lines are concrete enough to plan against. Horizon Europe’s bridging action “Bringing innovative climate preparedness and resilience solutions closer to the market” is identified as call HORIZON-BRIDGING-2027-04-03. The European Innovation Council Accelerator Challenge “Deep Tech for Climate Adaptation” is aimed at SMEs and start-ups, including nature-based and engineering approaches that use AI. The Communication also says the Commission will work with standardisation bodies on testing, performance, and interoperability of resilience products, and that it will explore a Climate Resilience Industry and Innovation Dialogue in 2027.

That is a consortia and product question as much as a compliance question. Our Horizon Europe consortium partnership is about productisation and adoption after the grant, not about relabelling an existing research outline as “resilience”.

The same investment logic is why cross-sector work such as The Adaptation Exchange exists: a shared way to assess resilience, a business case that finance can read, and examples of solutions that reduce real losses. The EU framework supplies a public timetable and a public vocabulary. It does not supply the asset-level case.

What is worth doing before the text is final

The proposal can still move. Delegated acts can later change the trajectory, the assessment checklist, the policy areas, and the indicators. A useful preparation does not depend on the final article numbers.

Treat 1.5°C / 2°C / 3°C as the baseline screen for sites, suppliers, and projects that will still be standing in 2050, and keep at least one worse case beside it.
Separate a regional hazard view (public EU tools) from an asset or catchment decision (local defences, elevation, vulnerability, who pays).
Write down who inside the organisation answers for each material risk, in language that can later map to a public risk owner.
For anything that will enter a permit, a tender, or a capital paper, keep the method note: sources, resolution, what was aggregated, and what would change the number.
Do not describe a vendor score, or a Copernicus map, as a residual-risk judgement. Residual risk is a decision about acceptable loss after measures.

The Commission has asked for a European approach to anticipate, prepare, and manage. The proposal is clearest on anticipation and on who should manage. Preparation of private and local investment still depends on evidence that survives a second look.

Sources

European Commission, COM(2026) 720, A Resilient Europe for a Changing Climate. Communication on the EU Framework for Climate Resilience, 27 October 2026. PDF
European Commission, COM(2026) 730, Proposal for a Regulation establishing a framework to enhance climate resilience (European Climate Resilience Law), 27 October 2026. PDF
European Commission, press release IP/26/2244, 27 October 2026, and news summary
EU Framework for Climate Resilience, Directorate-General for Climate Action
European Environment Agency, Navigator of Climate Resilience Tools, 27 October 2026
Eversheds Sutherland, EU: Climate Resilience Law proposal brings climate risk into public decisions, 2 November 2026
E3G, EU Climate Resilience Framework: laying foundations to address worsening climate impacts
World Resources Institute, statement on the EU Climate Resilience Framework
European Parliamentary Research Service, European integrated framework for climate resilience, briefing of 2 September 2026 (written before adoption of the proposal; useful for the prior political debate)