Cardano and Solana are putting their on-chain governance systems to the test in real time. On Cardano, a governance stalemate is looming because participation is too low, while Solana’s first major vote is running into a conflicting quorum rule.

On-chain governance allows token holders or validators to vote directly on protocol changes in a public process, without routing decisions through a core team. Cardano divides that responsibility among three separate bodies, while Solana assigns it to the delegated stake of its validators. Since the Voltaire era, Cardano has followed the CIP-1694 standard, under which most decisions require approval from at least two of the three bodies. Solana’s framework is much newer. It went live in July 2026 and requires 100,000 SOL in stake to submit a proposal. The two timelines also converge: Solana’s ballot closes on August 27, while Cardano’s replacement vote must take effect by September 1, making the coming days an unusually concentrated test of whether either design can convert participation into binding decisions. On Cardano, approval among DReps stands at 43.0%, while 67% is required. On Solana, 83.66 million SOL backed the most contested proposal, but whether that is enough depends on which rule is applied.

Cardano faces a governance blockade before the September 1 deadline

The proposal to renew Cardano’s Constitutional Committee requires two independent majorities. First, the delegated representatives, or DReps, must reach 67% approval. Second, the stake pool operators, or SPOs, must reach 51%. DReps are elected representatives to whom ADA holders assign voting rights. SPOs operate the pools that produce the network’s blocks. Each group must clear its own threshold, meaning strong support in one body cannot compensate for a shortfall in the other.

The trend has moved upward, but not fast enough so far. In mid-August, DReps stood at 32.46% and SPOs at 1.95%. Four days later, those figures had risen to 37.9% and 7.93%. In the latest reading, they had increased again to 51.7% and 10.8%. Even so, the gap remains large: roughly 15 percentage points for DReps and about 40 percentage points for SPOs. Time is running short, because the replacement vote must take effect on-chain by September 1, 2026. The nominees were already chosen in an off-chain election, and the on-chain vote has been underway since late July. The two-step design — an off-chain election to select nominees, followed by on-chain ratification — means the community has already settled on the candidates; what remains undecided is whether enough stake confirms them.

Four of the seven seats on the committee are expiring. If the vote fails, only three members would remain. The committee’s role is to review whether governance actions comply with the Cardano constitution. Under the reported rules, that requires at least five active seats. Even so, block production would continue normally and the network would still process transactions as usual. What would be affected are hard forks, protocol parameter changes, constitutional amendments and treasury payouts. Intersect, the membership organization behind Cardano’s development, has also warned of technical consequences. In its weekly update, it said a failed ratification could affect the timeline of the Dijkstra hard fork.

Solana’s quorum rule is internally inconsistent

Solana’s first voting package includes three proposals: SGP-0001 on the constitution, SGP-0002 on disinflation and SGP-0003 on fees. The voting window runs until the end of epoch 1023 on August 27, 2026. In the previous day’s snapshot, 83.66 million SOL voted in favor and 12.01 million voted against. Abstentions accounted for another 8.32 million SOL, bringing the total to roughly 104 million SOL. Among the decisive votes, that equals 87.45% approval.

The issue is not the size of the majority, but the quorum standard. Solana’s governance FAQ says turnout must reach one third of the network’s stake, with two thirds approval among participating votes. However, reports say the Solana Foundation’s proposal repository on GitHub sets no quorum at all and instead defines the threshold as two thirds of yes and no votes, excluding abstentions. Under the repository rule, SGP-0002 passes. Under the FAQ, turnout falls short of the one-third requirement. As a result, the first major test of the system ends with an unresolved interpretation dispute.

At its core, the proposal is about issuance. SGP-0002 would accelerate annual disinflation from 15% to 30%. Disinflation is the annual rate at which new SOL issuance declines, and the terminal value of 1.5% is the long-run floor the schedule converges toward. Because issuance determines how quickly new SOL enters circulation, the vote is ultimately about how fast total supply grows — and how much new supply is available as staking yield. Under that schedule, the network would reach the terminal value of 1.5% in about 2.8 years instead of 5.7 years. Over six years, that would reduce issuance by roughly 18.9 million SOL, or about USD 1.91 billion at the current price. Nothing changes immediately, because the proposal only sets the direction. Technical implementation would come later through SIMD-0550. In addition, 308 delegators overrode their validator’s vote for their own stake account, although that number is small relative to the overall vote.

Solana Company votes against faster disinflation

Solana Company holds SOL as a balance sheet reserve and also runs its own validators. The Nasdaq-listed company, HSDT, disclosed its positions before the voting window opened. It voted against SGP-0002 and SGP-0003, and in favor of SGP-0001. That means one of the most visible institutional holders is opposing faster disinflation. Under stake-weighted voting, a reserve of that size carries proportionate weight in the tally.

The company’s own financials help explain why. In the second quarter of 2026, USD 2.512 million of its USD 2.526 million in total revenue came from staking, or about 99.4% of quarterly revenue. Faster disinflation reduces issuance and therefore lowers staking yield. Solana Company is, in effect, voting on a rule that directly affects its main source of revenue. That is the structural conflict of interest built into stake-weighted governance: the largest stake holders often have the biggest economic interest in maintaining the status quo.

The company says its opposition is about process, not the goal of lower issuance. It argues that reopening an already deterministically fixed issuance schedule would introduce uncertainty. It also says staking yield is an audited balance sheet item and, for many holders, operating cash flow. Solana Company laid out those positions in a statement.

“We strongly believe that institutional adoption is a key driver of Solana's growth, and institutions make decisions based on consistent, predictable structures.” - Joseph Chee, Chairman and CEO, Solana Company

Two governance models, one shared weakness

Cardano and Solana use very different governance structures. On Cardano, DReps and SPOs vote by stake, while the Constitutional Committee votes by head count. Two of the three bodies must agree, which is designed to encourage broad participation. On Solana, the decision is concentrated among validators, who by default vote with the stake delegated to them. Individual stakers can still override their validator’s choice for each stake account, and that portion is then removed from the validator tally.

Neither system is failing because of the technology itself. Cardano’s problem is lack of participation, even though the rules have been known for months and the deadline is clear. Solana has enough votes to matter, but lacks a clearly defined rule for determining when those votes are sufficient. The two problems are familiar from traditional shareholder democracy as well: voter apathy at the base, and large voting blocs with their own interests. The immediate checkpoints are already fixed: Solana’s ballot closes at the end of epoch 1023 on August 27, Cardano’s ratification must be on-chain by September 1, and any change to Solana’s issuance would still require the later SIMD-0550 implementation. The packaging is new, but the underlying issue is not.