CORSIA's first phase ends on 2026-12-31 and ICAO counted 40,650,719 tonnes of eligible emissions units issued and labelled as of 2026-07-28 — units from Council-approved programmes that must meet double-claiming criteria such as a host-country letter of authorization, while Assembly Resolution A42-22 assigns no body responsibility for how many units exist

What is happening?

CORSIA is the market-based measure of the International Civil Aviation Organization that requires growth in carbon dioxide emissions from international flights to be offset. In 2025 the 42nd ICAO Assembly reaffirmed it, in Resolution A42-22, as the only global market-based measure applying to international aviation emissions. Its first phase runs from 2024 to 2026 and ends on 2026-12-31. Aircraft operators then have to cancel eligible emissions units to cover the obligation, with the cancellation deadline given in opened sources as December 2027 or January 2028 depending on the source.

The rule is in force and the obligation is running. What lags is the input the rule demands. An eligible unit only exists after a chain of separate decisions: the ICAO Council approves a crediting programme on the advice of its Technical Advisory Body, a host country takes action against double claiming, for example by issuing a letter of authorization, and a programme registry issues the unit and labels it as eligible. As of 2026-07-28 the ICAO dashboard counted 40,650,719 tonnes of units that had been issued and labelled this way, from 10 host countries.

How many units the first phase will need is not stated in any ICAO document opened here. Estimates from outside ICAO run from 100 to 150 million tonnes, attributed to ICAO by an environmental nonprofit, to 170 to 236 million tonnes in an airline industry association paper. Every one of those figures is larger than the dashboard total. Resolution A42-22, read in full, gives the Council the job of approving programmes and gives member states the job of building domestic implementation frameworks, and contains no paragraph that makes any body responsible for the quantity of units available.

Whose problem is this?

Operator counts differ between two ICAO documents: 690 in the 2025 reporting cycle according to a dashboard capture in a 2026-06 regional presentation, and 675 in the 2026-07 newsletter while submissions were still in progress.

Where does this problem end?

The boundary sits on the chain rather than on any single actor because each link is held by a different body and none of them is breaking a rule by moving slowly.

What is the state now, and what should it be?

Now

Needs a new measurementthe target state: no ICAO document opened here states how many eligible units the first phase requires, nor sets any target for unit supply by a given date. ICAO has asked states to issue authorization letters expeditiously, in its 2025-07-25 letter and again in a 2026-06 regional conclusion, but that is a request about speed and carries no quantity. The demand figures in this dossier come from outside ICAO and are not used as a target here.

How big is it?

The measured quantity is supply already issued and labelled: 40,650,719 tonnes as of 2026-07-28. Every other quantity in this problem is an estimate from a non-ICAO source and they do not share a basis.

On demand, an airline industry association paper presented at a 2026-09 Asia Pacific meeting of civil aviation directors general put first-phase demand at 170 to 236 million tonnes, using the 2024 sector growth factor, and its 2026-09-18 supply chart said 200 to 250 million tonnes would be needed by January 2028. An environmental nonprofit post dated 2026-07-27 cited an ICAO estimate of 100 to 150 million tonnes, a figure it dates to 2025-07, and 880 to 1,350 million tonnes for the second phase.

On supply, the same association counted about 38 million tonnes usable for the first phase as of 2026-06, and about 72 million tonnes covered by 42 authorization letters as of 2026-08, describing a gap of about 150 million tonnes. The 72 million tonne figure includes verified volume not yet issued, so it is a different quantity from the ICAO dashboard total and the two should not be subtracted from each other.

Under what conditions does it arise?

1. Each link has a different owner. The Council approves programmes, a host-country government decides on authorization and the corresponding adjustment, and a registry issues and labels units. A unit is only usable when all three have acted. No link can compel the next one.

2. The host country carries a cost for authorizing. Authorizing a unit for international use means the host country makes a corresponding adjustment under the Paris Agreement, so the reduction is not also counted toward its own target. The industry association attributes slow issuance to uncertainty about corresponding adjustments and low awareness among finance ministries. That is the association reading, not a position stated by any host government.

3. The obligation has a calendar and the supply chain does not. The first phase ends on 2026-12-31 and cancellation follows by early 2028. Nothing in the chain is tied to that calendar. Supply is also concentrated, with about 61 percent of labelled tonnes from one host country as of 2026-07-28, so decisions by a small number of governments move most of the total.

4. The alternative compliance route is small. Operators can also reduce obligations with eligible fuels, but the dashboard records 181,263 tonnes of certified eligible fuel from 2024 production, and the industry association expects first-phase obligations to be met mainly with units.

What has been tried?

Two of these act on the chain from the demand side and from the advisory side, and neither has authority over the link that host governments hold. That is why every attempt in this table takes the form of a request, an offer of help or a review.

What was found?

Why is it still unsolved?

Coordination failure — every link in the supply chain belongs to a different body, each acts within its own rules, and nothing ties their combined output to the date when operators must cancel units.

The Council can declare a programme eligible but cannot make a host country authorize anything from it. A host government can authorize units but carries the corresponding adjustment when it does, and that cost lands in its own national accounts rather than in the aviation scheme. Registries issue and label what has been authorized and nothing more. The operators that need the units hold none of these levers.

Resolution A42-22 divides the work exactly along those lines. It asks the Council to approve programmes, asks states to build domestic frameworks, asks ICAO and states to support capacity building, and asks the Council to facilitate the use of units that benefit developing states. None of those paragraphs makes anyone answerable for whether the total is enough. The periodic review exists, but the opened sources do not show it being applied to first-phase supply.

The industry association attributes slow authorization to uncertainty about corresponding adjustments and low awareness in finance ministries. That diagnosis comes from a party with a direct stake on the demand side, and no host government statement was found to confirm or contest it.

What observation would mean it is solved?

Candidates — (a) issued and labelled supply on the ICAO dashboard reaches a published estimate of first-phase demand before the cancellation deadline (b) ICAO publishes its own first-phase demand estimate alongside the supply dashboard, so the gap is stated on one basis (c) a document assigns a body responsibility for monitoring unit quantity against the compliance calendar.

(a) alone depends on which demand figure is chosen. The estimates opened here range from 100 to 236 million tonnes, so the same supply total could look sufficient or less than half of what is needed.

(b) alone measures the gap without closing it. A published ICAO estimate would end the dispute over the basis, but it would not oblige any host government to authorize anything.

(c) alone can exist on paper. A monitoring responsibility with no authority over host-country decisions would describe the chain without changing its speed. Read together, the three would show a measured gap, an owner, and a total moving toward it.

What is it connected to?

Fills with researchthe Paris Agreement Article 6 authorization and corresponding adjustment process in general, voluntary carbon market supply, the second CORSIA phase from 2027, eligible aviation fuel production, and the related corpus entries on shipping carbon pricing and the loss and damage fund. Relation type and evidence grade were not confirmed in this round.

What these sources do not say

• ICAO own first-phase demand estimate. No ICAO document opened here contains one. A search result pointed to an ICAO file on supply and demand projections, but it returned HTTP 403 and was not read. The sources do not say ICAO made no estimate; the opened range simply contains none.
• What happens if supply falls short. A42-22 sets out safeguards in paragraph 16 and periodic review in paragraphs 9(g) and 17, but no paragraph states an effect of a first-phase supply shortfall. The compliance provisions of Annex 16 Volume IV were not opened, so this is not a finding that such provisions are absent.
• Whether any other ICAO document assigns responsibility for quantity. A42-22 does not. Annex 16 Volume IV and the eligibility criteria documents were not opened.
• The exact wording of the double-claiming criteria. The dashboard note gives a letter of authorization as an example; the criteria document itself was not read.
• Whether ICAO has reconciled the differing demand and supply bases of the industry association figures. No source addresses it.
• The position of host-country governments. The reasons for slow authorization come from the industry association. A direct statement from any host government was searched for and not found.
• Sanctions for first-phase non-compliance by state. No source opened here describes them beyond a mention of domestic legislation in a law firm market note.
• Which date is the cancellation deadline. Opened sources give December 2027 and 2028-01-31, and none of them reconciles the two.

See the evidence

One primary ICAO decision text was read in full, Resolution A42-22. The ICAO newsletter, regional presentation and eligible units page were opened directly and carry the dashboard figures, programme counts and the authorization requests. Everything about demand, and the alternative supply counts, comes from an airline industry association, an environmental nonprofit, a law firm note and trade press. Those sources do not share a basis — the industry supply count includes verified but unissued volume, the dashboard does not, and the demand estimates range from 100 to 236 million tonnes — and the disagreement is left visible rather than resolved. Annex 16 Volume IV, which holds the compliance provisions, was not opened. This is a Path A output (research-based definition), so observation_refs is empty and provenance_mode: press-derived.

People affected

Estimated range Not derivable

The reason and what is missing are listed under “What is missing” below