A group of six Ethereum researchers has proposed a change that would reduce validator rewards as more ETH is staked, eventually burning all newly issued staking rewards if staked ETH reaches about 60.25 million ETH, or roughly half of the total supply. The draft proposal, known as EIP-8361, would not redirect those rewards elsewhere; it would destroy the newly created ETH instead.
The proposal matters because it targets Ethereum's monetary policy and staking incentives at the same time. Its authors argue that Ethereum's current reward curve keeps encouraging more staking even at high participation levels, which could push more ETH toward large exchanges and staking providers and weaken decentralization.
Under the plan described by CoinDesk, the burn would increase gradually as the staking ratio rises. Validators would continue receiving transaction fees and tips from block production, while only newly issued ETH rewards would be affected. The change would phase in over roughly 18 months, with an estimated six-month lead time before activation, giving the network about two years to adjust.
The debate comes as about 41 million ETH, close to 34% of supply, is already staked, with another 2.5 million ETH waiting in the validator activation queue. One of the proposal's authors, Jerome de Tychey, projects that staked ETH could exceed 70 million by January 2028 if the current system remains unchanged.
The proposal has split Ethereum developers and DeFi participants. Aave Labs CEO Stani Kulechov warned that lower staking rewards could weaken ETH borrowing strategies, while ether.fi founder Mike Silagadze criticized both the process and the potential impact on liquid staking and DeFi capital flows.
It is unclear whether EIP-8361 will make it into Ethereum's Hegota upgrade, which is planned for the second half of 2026 and is expected to focus on structural cleanup, censorship resistance and state size reduction. CoinDesk reported that the proposal arrived shortly before the upgrade's inclusion deadline, making a later fork more likely if consensus does not form quickly.