Four empty seats could freeze Cardano’s treasury on 1 September

• Four of the seven seats on Cardano’s Constitutional Committee expire at epoch 653, which lands around 1 September. Drop below five members and most governance actions can no longer pass.
• What would stop: treasury spending, protocol parameter changes, hard forks. What would not: your transactions, your staking rewards, block production, every app running on the chain.
• The four replacements have already been elected, but with days left the renewal vote is well short. Delegated representatives were at 41.7% against a 67% threshold on 25 August, and stake pool operators at 12.0% against 51%.

Cardano runs on a written constitution and an elected committee whose job is to check that new decisions comply with it. Four of that committee’s seven seats expire in the first days of September, and with a week to go the vote to fill them is a long way short of the support it needs. If it doesn’t pass, the chain carries on producing blocks and most of its governance quietly stops working.

What Cardano voted itself into

For most of its life Cardano was steered, in practice, by a small number of organisations that built it. In 2024 and 2025 it did something unusual and handed those decisions to on-chain voting, with a written constitution sitting on top. Three groups now share the work, and none of them can act alone.

The first group is delegated representatives, always shortened to DReps. Cardano’s staking system already lets you delegate your coins to a pool without giving up ownership, and governance copies the idea: you delegate your voting power to a DRep, who then votes on proposals using the weight of the stake behind them. You can vote yourself if you want to register as your own DRep. Most people don’t.

The second group is stake pool operators, or SPOs. These are the people running the machines that actually produce blocks. They vote on a narrower set of questions, the ones that touch the security and structure of the network rather than everyday spending.

The third is the Constitutional Committee. Seven seats, and a job that is easy to misread.

The committee isn’t a board, it’s a check

The committee doesn’t decide what Cardano should do. It can’t propose spending and it can’t originate policy. Its role is narrower: when a proposal is put on-chain, the committee votes on whether that proposal is constitutional. Think of it less as a cabinet and more as a compliance gate.

Because it’s meant to be a check, it deliberately can’t defend itself. The committee does not vote on motions of no confidence in itself, and it does not vote on its own renewal. That’s why the four elected replacements can’t simply be waved through by the three members staying on. A committee that could veto its own replacement wouldn’t be a check on anything.

The same logic produces the minimum size. Cardano’s rules require at least five sitting members for the committee to function, so that constitutionality is judged by a group rather than by whoever happens to be left. Lose four of seven and three remain, which is below the floor.

The deadline is an epoch boundary, not a date someone chose

Cardano measures time in epochs: fixed windows of five days used for scheduling staking rewards and, now, governance terms. The four seats expire at epoch 653, which falls in the first days of September.

That matters because an epoch boundary is not a deadline anyone can negotiate. There’s no grace period written into the rules and no committee that can extend its own term while the paperwork catches up. The renewal action went on-chain on 31 July, after Cardano completed and independently audited its 2026 committee election. Either enough voting weight arrives before the boundary or it doesn’t. With a week left, it hadn’t.

What actually stops, and what carries on

Intersect, the member-based organisation that coordinates much of Cardano’s development, spelled the consequence out on 14 August. If the committee falls below its minimum, only two kinds of governance action remain available: Info Actions and Update Committee actions.

An Info Action is an on-chain poll. It records what voters think and does nothing else, with no automatic effect on the chain. An Update Committee action is the one that changes who sits on the committee, which is the escape hatch: a lapsed committee can still be repaired by the same mechanism that was supposed to renew it.

Everything else waits. Treasury withdrawals, the payments that fund development work, can’t be approved. Protocol parameter changes can’t be approved either, which covers things like fees, block size and deposit requirements. Hard fork initiation, the route to any upgrade requiring new node software, is off the table. So is ratifying a new constitution.

Now the part that gets lost in headlines about a chain freezing. The blockchain itself is unaffected. Blocks are produced on the normal schedule, transactions settle, staking rewards accrue, and every application, exchange listing and DeFi pool built on Cardano works exactly as it did the day before. Nothing about holding or moving ADA changes. What freezes is the committee room, not the ledger.

It’s worth being clear that this is roughly the same state Cardano would enter deliberately if voters ever passed a motion of no confidence, which dissolves the committee outright. The restricted mode isn’t a bug someone forgot about. It’s the designed consequence of having no functioning constitutional check, and the design assumes the gap gets closed quickly.

Why a vote like this can fail without anyone opposing it

The obvious assumption is that a failed renewal means voters rejected the candidates. That’s not usually how these things go wrong.

Cardano’s thresholds are measured against voting stake, and stake that doesn’t turn up counts for nothing. ADA that has never been delegated to a DRep is silent by default, and a DRep who doesn’t vote leaves that weight on the table. A proposal can be broadly uncontroversial and still miss its bar simply because not enough delegated stake showed up in time.

The tallies now show that happening. A CardanoScan reading reported by CryptoSlate on 25 August put DRep support at 41.7%, against the 67% the action needs. Stake pool operators were at 12.0%, against a 51% bar. Both groups have to clear their own threshold for the same action, and the SPO gap is by far the wider of the two.

Support is rising, and it has been rising throughout. A GovTool snapshot on 17 August had DReps at 32.46% and SPOs at 1.95%. Intersect’s own reading on 21 August put them at 37.9% and 7.93%. That is roughly nine points of DRep support and ten points of SPO support added over eight days. Our own arithmetic, rather than anyone’s forecast: keep that pace and the DRep bar is still missed and the SPO bar is nowhere near, so the vote now needs turnout to accelerate sharply in its final days rather than continue as it has been.

There is recent precedent. In June, a treasury vote to fund Cardano’s Singapore summit missed its threshold and the event was cancelled, an outcome that turned out to be about participation rather than any argument over whether the summit was worth having. That was an event. This is the body that signs off on everything else.

What to watch

With days rather than weeks left, three things tell you where this ends up.

First, the SPO number. It is the one that decides this, because it needs to more than quadruple before the epoch boundary while the DRep figure needs to add around 25 points. Stake pool operators are a smaller and more coordinated group than the DRep base, so a concerted push can move that percentage quickly. If it stalls in the teens, the seats lapse.

Second, whether idle voting power moves. If a meaningful amount of undelegated stake gets delegated to DReps in the final days, the arithmetic changes fast. If it doesn’t, the outcome rests entirely on the DReps and operators already active.

Third, and now the likelier scenario, how quickly an Update Committee action goes up and passes if the seats do expire. Because that action type stays available in restricted mode, a lapse is recoverable rather than terminal, and the length of the gap will say a great deal more about Cardano’s governance than the lapse itself. A fortnight of restricted mode is an administrative embarrassment. Months of it, with treasury payments stalled and no route to a parameter change, would be a different story, and one worth reading alongside the rest of our policy and governance coverage.