Last verified 12 August 2026. Four items carry status risk and are flagged where they appear: the EPA endangerment rescission is under consolidated challenge in the DC Circuit, confirmed unstayed only to mid-April 2026; the vacatur of IRS Notice 2025-42 in Oregon Environmental Council v. IRS is confirmed to 6 June 2026 and is subject to appeal and to reissued guidance on remand; the Industrial Accelerator Act, the CBAM downstream extension and the 90% 2035 vehicle target are proposals rather than law; and the 1.2% maritime target in Directive (EU) 2023/2413 is recital language creating no duty.
In July 2025 a reconciliation law rewrote the economics of American cleantech. In February 2026 the EPA withdrew the legal basis for regulating vehicle emissions. Much of the demand those policies had created moved to Europe.
Over the same period the EU built two mechanisms that do what US policy used to do: make the polluting option more expensive, and put a legal obligation behind the clean one. For companies in the right sectors, the customer that federal policy stopped creating at home now exists in Europe by law.
American capital remains deeper and faster. The change is in where public policy creates a buyer, which makes this a question about where a company puts its weight. A Canadian, Australian or European company choosing which market to build around runs the same two tests as an American one.
Underneath the detail, both blocs now push production onshore. The US does it through ownership restrictions on its surviving manufacturing credit and through domestic preference terms attached to federal awards. Europe does it by scoring procurement on where a product was made and by requiring real operations in a member state before it funds anything. For most companies the open question is which market to localise in first.
Where the sectors land
The 2025 law hit cleantech unevenly. Hydrogen lost most of its support, carbon capture gained, and several sectors were left alone.
Hydrogen and e-fuels
- US: the section 45V clean hydrogen production credit, worth up to $3.185 per kilogram in 2025, being the $3.00 statutory amount adjusted by the 1.0611 inflation factor in Notice 2025-37, for producers meeting prevailing wage and apprenticeship requirements, had its construction deadline pulled forward from 2033 to 2028 by the July 2025 reconciliation law
- Europe, more favourable: FuelEU Maritime, Article 4 of Regulation (EU) 2023/1805, sets a binding greenhouse gas intensity trajectory for ships, applying since 1 January 2025. A 1.2% share of renewable fuels of non-biological origin in maritime energy appears in Recital 72 of Directive (EU) 2023/2413, in the preamble, phrased as an endeavour and creating no enforceable duty. Article 4 carries this row on its own, and it is the strongest case on this list
Clean fuels
- US: the section 45Z clean fuel production credit was extended by two years, to fuel sold through 31 December 2029, though the same law cut the sustainable aviation fuel rate from $1.75 to $1.00 per gallon for fuel produced after 31 December 2025 and loosened lifecycle rules in favour of conventional biofuels, diluting the advantage for synthetic producers
- Europe, more favourable: the same FuelEU Maritime and Renewable Energy Directive obligations, reinforced by full inclusion of shipping in the EU carbon market from this year
Solar and wind
- US: sections 45Y and 48E credits end for wind and solar placed in service after 31 December 2027 unless construction began by 4 July 2026. The standard for proving construction began is currently in dispute, covered below
- Europe, more favourable: the Net-Zero Industry Act requires member states to apply non-price criteria, including a supply-resilience criterion, to at least 30% of volume auctioned per year per member state, or alternatively at least 6 GW per year per member state, under Article 26 of Regulation (EU) 2024/1735, applying since 30 December 2025
Manufacturing
- US: the advanced manufacturing production credit survives, with new restrictions tied to foreign ownership. Manufacturers have reason to stay
- Europe, equal: the same Net-Zero Industry Act criteria reward production inside Europe and count against an exporter, so the advantage arrives only with establishment. Comparable rules apply to public procurement of net-zero technologies
Carbon capture
- US, improved: the 45Q credit was preserved and improved, with $85 per tonne for point-source capture and $180 for direct air capture holding unadjusted through 2026 and inflating from 2027, and utilisation raised to parity with storage for equipment placed in service after 4 July 2025
- Europe, less favourable: no EU instrument obliges anyone to buy captured CO2, and the Carbon Border Adjustment Mechanism leaves out methanol and most finished chemicals. A second market, with no urgency
EVs, charging, efficiency
- US: consumer and commercial credits eliminated, and the EPA rescinded the greenhouse gas endangerment finding and motor vehicle greenhouse gas standards in a final rule published 18 February 2026, 91 FR 7686, effective 20 April 2026, now under consolidated challenge in the DC Circuit with no stay granted as at mid-April 2026
- Europe, more favourable but weakening: the 100% carbon dioxide reduction for new cars from 2035 under Regulation (EU) 2019/631 remains the adopted requirement. The Commission has proposed cutting it to 90% in COM(2025) 995 final, tabled 16 December 2025, with a plenary vote expected 23 November 2026. That is a proposal and not law
Storage, nuclear, geothermal
- US: credits untouched until 2034
- Europe, less favourable: no purchase obligation, and no scoring advantage that turns on production location. Ordinary market entry
Three of these need more detail. All three are about the US.
US: hydrogen and e-fuels lost the most. The clean hydrogen credit paid up to $3.185 per kilogram in 2025. Only producers meeting prevailing wage and apprenticeship rules got that top rate. Hydrogen is the largest input cost in e-fuels and e-chemicals, so a subsidy that size carries the business case on its own. Clean Air Task Force reported the effect straight away: aviation fuel producers doubting they could continue, and hydrogen projects worth billions paused or abandoned.
US: carbon capture improved. The credit held at $85 per tonne for capture at a plant and $180 for capture from the air. Using the CO2 now earns the same as storing it. FactSet reads this as opening new revenue from selling CO2 into products, including e-fuels. So the same law helps a company that captures CO2 and hurts one that converts it, because converting needs hydrogen. That company gets cheaper feedstock and loses the subsidy on its main process.
US: solar and wind have a deadline that has passed, and a contested rule behind it. Projects not running by 31 December 2027 had to start construction by 4 July 2026 to keep the credits. What counts as starting construction is now in dispute. Treasury tightened the test in Notice 2025-42 in August 2025, requiring physical work and withdrawing the option of spending 5% of project cost, except for solar facilities of 1.5 MW or less. On 6 June 2026 the US District Court for the District of Columbia vacated that notice as arbitrary and capricious in Oregon Environmental Council v. Internal Revenue Service, No. 25-4400 (CKK), and remanded it, restoring the 5% option. That is the position as at 12 August 2026. An appeal to the DC Circuit, a stay, or reissued guidance on remand would change it. Any project relying on the 5% route should treat its eligibility as provisional and take tax advice.
What Europe built
Europe is not the stable jurisdiction, and any article claiming otherwise is worth closing. The EU narrowed the scope of its sustainability reporting rules in Directive (EU) 2026/470, adopted 24 February 2026. It has proposed cutting the 2035 requirement for new cars from a 100% carbon dioxide reduction to 90% in COM(2025) 995 final, which remains a proposal. Both blocs have trimmed their ambitions. What separates them is that Europe built two mechanisms that create a buyer.
One: a mandated buyer for molecules
This has no US equivalent, which is why companies used to American policy design tend to miss it.
FuelEU Maritime sets a binding trajectory for the greenhouse gas intensity of fuel used by ships: 2% below the 2020 baseline by 2025, 6% by 2030, 14.5% by 2035, rising to 80% by 2050. Renewable fuels of non-biological origin, the category that covers e-methanol, e-ammonia and e-hydrogen, have their intensity halved in the calculation from 1 January 2025 until 31 December 2033, so a tonne counts twice. A harder rule sits behind it: if RFNBOs are less than 1% of in-scope fuel in 2031, a 2% RFNBO target applies from 1 January 2034. Underneath it, the EU carbon market now covers shipping emissions in full, phased in from 40% in 2024 to 100% this year.
European buyers of marine fuel therefore face a rising carbon cost on the fossil option and a legal obligation on the clean one.
Two: a scored advantage for hardware
The second mechanism applies to equipment.
The Net-Zero Industry Act criteria include a resilience contribution, which measures how concentrated a supply source is. Only a producer established in Europe can score on it. The criterion applies where the Commission has found that a single non-EU country supplies more than 50% of EU annual demand for a net-zero final product or its main specific components, or more than 40% where that share rose by at least 10 percentage points on average across two consecutive years. Contracting authorities must use the Commission's published dependency communication and cannot calculate dependency themselves. Guidance on Article 25 and Article 26 was published on 22 July 2026 as C/2026/5024 and C/2026/4908.
Separately, the Carbon Border Adjustment Mechanism entered its definitive phase on 1 January 2026 under Regulation (EU) 2023/956, covering imported cement, iron and steel, aluminium, fertilisers, electricity and hydrogen. Regulation (EU) 2025/2083, adopted 8 October 2025, added a de minimis of 50 tonnes per importer per calendar year across iron and steel, aluminium, fertilisers and cement, and deferred the first declaration and certificate surrender for 2026 imports to 2027, with certificate sales opening 1 February 2027. No importer's cash position changes during 2026. Scope leaves out methanol and most finished chemicals. A proposed extension to roughly 180 downstream goods, COM(2025) 989 final of 17 December 2025, would apply from 1 January 2028 and remains a proposal.
A third instrument, the Industrial Accelerator Act, was tabled in March 2026 and would link EU support schemes to Union-origin criteria. It remains a proposal, weaker as tabled than earlier drafts, and is best read as direction of travel.
Two honest qualifications. The resilience criteria were written with Chinese supply concentration in mind, and the mechanism makes no distinction by nationality, so a US exporter sits on the wrong side of it either way. And none of this makes European cleantech more profitable in general. It makes production location a scored variable in specific purchasing decisions, which is the narrower claim that holds up.
What that is worth
For a producer, these instruments hold up the price a clean product can command.
Fossil fuel now carries a carbon cost. With shipping fully inside the carbon market this year, conventional marine fuel at roughly 3.2 tonnes of CO2 per tonne of fuel picks up around 320 euro per tonne at a carbon price near 100 euro. Buyers compare a clean fuel against fossil plus carbon.
The product counts twice. The double-counting multiplier for renewable fuels of non-biological origin runs to the end of 2033. One tonne delivers two tonnes of compliance value to the buyer, which lifts what they can rationally pay. It expires, which is why early offtake conversations are worth more than later ones.
Missing the target costs money. Falling short of the intensity trajectory carries financial consequences for the shipping company under Regulation (EU) 2023/1805. Separately, Article 6 requires containerships and passenger ships at quayside to use onshore power supply or certified zero-emission technology, from 1 January 2030 in EU ports covered by Regulation (EU) 2023/1804 and from 1 January 2035 in other EU ports with onshore power installed. A buyer who cannot source compliant fuel has no cheap fossil fallback.
For hardware the equivalent is narrower but real: a protected share of auction volume that only an established producer can compete for.
None of this guarantees margin. Legislation, with dates attached, sets the ceiling for a compliant product above the fossil reference.
Which test applies
The demand test. Does European regulation oblige someone to buy the product, or penalise them for buying the incumbent? This catches marine and aviation fuels, renewable hydrogen, and qualifying e-chemicals.
The location test. Could a European customer be required to score a bid on where the product was made, or will the goods carry a carbon border cost? This catches solar, batteries, storage hardware, and producers of the covered industrial goods.
Both tests come out weak for companies selling to private buyers with no procurement criteria, for software, and for anything with short commercialisation cycles that can absorb a policy shift.
Europe subsidises the entry
The EIC Accelerator funds single companies scaling breakthrough technology: a grant of up to 2.5 million euro, plus an investment component between 1 million and 10 million.
For a company at scale-up stage, 12.5 million euro is on its own a thin reason to restructure, and American rounds routinely exceed it. The grant comes once per company, and the investment component runs through separate due diligence that can defer or decline it.
It is better understood as a subsidised route to building the European position these rules increasingly require. Where a European entity is needed to reach a mandated market, the EIC pays for a meaningful share of standing it up, non-dilutively on the grant side.
A company from a non-associated third country can submit a short proposal, but must prove effective establishment in a member state or associated country by the full proposal stage.
These two requirements point the same way, which is the useful part. Effective establishment means running real operations in Europe, and the grant funds exactly those operations: the European pilot, the demonstration line, the regulatory and certification work. The same footprint that satisfies the EIC also puts production inside Europe, where the Net-Zero Industry Act criteria can score it. Public money pays for the step that makes a later tender bid credible.
The sequence has three stages rather than one. The grant covers innovation activities up to TRL 8. Commercial production capacity falls to the investment component, and first-of-a-kind capex to instruments such as the EU Innovation Fund. What the grant buys is the European foothold and the reference plant that the rest is financed against.
One point for applicants controlled from outside the EU. Some grant-only calls restrict such beneficiaries, and carve-outs vary. The restriction is call-specific, and the eligibility text in the relevant Challenge of the EIC Work Programme 2026 should be read directly before it is relied on.
The trade, stated honestly
The part most founders outside Europe get wrong comes first, because it ends conversations before they begin.
Existing ownership is retained. Background, meaning the knowledge and rights that existed before the project and are needed to run or exploit it, stays with its owner. A non-EU parent licenses its patents to the European entity with the rights that entity needs, and continues to own and use them everywhere else.
What the grant pays to create sits in Europe. Results belong to the beneficiary that generates them. The European entity is the beneficiary, so the results developed under the grant, known as foreground, are owned there. Foreground carries an obligation to exploit, along with constraints on transferring it or granting exclusive licences to parties outside the EU and associated countries.
The commitment therefore applies only to what comes next. The existing IP position stays where it is, and what changes is that the next layer, the one European public money paid for, is owned by and exploited from the European entity.
That is still a real decision. Work funded in Europe compounds into a European asset, and moving it out later is subject to constraints. It is a choice about where a company builds its second base.
What is on offer is subsidised entry to a market that legislates what its buyers must purchase. For companies whose home subsidy just moved, that trade looks better than it did two years ago.
Where to start
Run both tests. If neither applies, Europe is an ordinary market entry question.
If one does, sequence matters more than speed. The entity has to exist with genuine operations before the full proposal, which takes months.
On IP, two things come before any structuring. Map what is background, meaning what is already owned and will be licensed in, against what will be generated under the grant and owned by the European entity. And where public awards are already held, their exploitation and manufacturing conditions come first. US federal funding is the common case, carrying domestic preference terms that constrain what can be licensed or manufactured abroad, and national programmes elsewhere can attach conditions of their own. That has to be understood before a European structure is built around it.
Inneuvate works through that sequence with companies based outside the EU regularly, usually well before anything is committed.
FAQ
Which sectors lost the most? Hydrogen, and solar and wind. The section 45V construction deadline moved from 2033 to 2028. Wind and solar credits end for projects placed in service after 31 December 2027 unless construction began by 4 July 2026, and the rule for proving construction began is currently subject to litigation.
Did anything improve? Yes. Carbon capture. The credit was preserved and utilisation raised to parity with storage at $85 per tonne, with direct air capture at $180. Storage, fuel cells, geothermal and nuclear keep their credits until 2034.
Does a non-EU parent have to hand over its intellectual property? No. What was owned before stays with the parent and gets licensed to the European entity. What the grant funds is owned by that entity, with obligations to exploit it and constraints on moving it outside the EU.
Where does new IP live if the engineers sit outside the EU? The grant funds costs incurred by the beneficiary, so the work has to happen in the European entity for those costs to be eligible. That is the substance behind effective establishment, and why the entity has to run real operations.
Is Europe more stable? Not straightforwardly. The EU narrowed its sustainability reporting scope in February 2026 and has proposed, but not adopted, a softening of the 2035 vehicle requirement. The difference lies in European rules increasingly obliging buyers to purchase clean molecules, or to score where a product was made.
Is the funding worth restructuring for? On its own, usually not. It is worth considering as subsidised entry to a market a company had commercial reasons to enter, and the case is strongest where home-market support has been withdrawn.