Is ETH in Danger of Becoming Too Centralized?
ETH was built on the promise of decentralization—no gatekeepers, no authority, just an open network driven by its people. However, as the , concerns are increasing that the second-largest blockchain is heading towards centralization. From the hegemony of a few staking providers to increasing reliance on infrastructure controlled by the large players, some question if ETH is staying true to its mission or if it is actually in danger of becoming too centralized.
Key Takeaways
- Liquid staking providers, including Lido, control a large share of staked ETH, creating economic and regulatory vulnerabilities.
- Heavy reliance on a few client software and MEV relays increases the chances of systemic failures and censorship.
- Centralized sequencers and compliance-driven transaction filtering erode ETH’s core principle of decentralization.
How Proof-of-Stake Contributes to Centralization
The event of 2022, which shifted the ETH network from to Proof-of-Stake (PoS), has redefined its security profile and the nature of centralization risks. Under PoS, network integrity is no longer tied to accumulating immense computational resources, but rather to controlling a significant economic stake.
The change made staking concentration an obvious place to watch: if too much stake is held by a handful of actors, it could influence consensus decisions, censor transactions, or cause outages that harm the network.
Economic Concentration and Staking Pool Dominance
The greatest external threat to ETH decentralization is the concentration of staked Ether (ETH) in a select few entities, primarily Liquid Staking Providers (LSPs). LSPs enable capital to be staked without meeting the 32 ETH requirement to run a Block confirmer, greatly enhancing convenience and liquidity. This act encourages centralization.
For instance, Lido Finance, a top liquid staking protocol, currently controls approximately 30.49% of the staked ETH. On its own, it sits near the 33.33% economically decisive threshold. With a few additional percentage points of market share, Lido would theoretically have economic power to halt finalization through coordinated action or default.
Staking Market Supremacy
LSP governance structure and operation decisions, along with centralized platforms (CEXs) for managing large staking pools, focus on controlling the majority of staked capital in comparatively small numbers of organizations. This structural arrangement suggests that the greatest threat is not necessarily a malicious internal attack, but rather external regulatory pressure leading to coerced collusion. If the larger LSPs or CEXs are mandated, 30% or more of the network that they represent can be compelled to comply, turning regulatory exposure into a systemic, protocol-level censorship risk.
The development community has explored implementing staking caps to mitigate this concentration, proposing an upper limit for delegators’ funds that would be significantly lower than previous staked totals. Decentralization intervention at the protocol level would similarly have the effect of suppressing the income and growth potential for staking service providers, pitting the network’s long-term integrity against short-term economic incentives for large players in tension.
Technical Monoculture and Client Risk
In addition to economic control, technical resiliency is eroded through the lack of diversity in the software used by network nodes. Technical monoculture is established when the majority of Block confirmers use a single client software. If that predominant client were harboring a fatal bug, it could crash or disrupt the network, resulting in an immediate, centralized, coordinated recovery procedure.
On the Execution Layer, Geth remains the clear supermajority client with a claimed market share of 62.62% according to Ethernodes statistics. While still short of the 66% mark required for a poor actor to finalize blocks, it is well above the secure limit for a non-malicious failure. A bug in Geth could disengage nahead two-thirds of the processing capacity, thereby severely disrupting the network. This technological vulnerability is a near-term and immediate threat because of a coding error rather than a high-level economic conspiracy.
However, measuring diversity on the Consensus Layer is challenging and varies across monitoring services. For example, some pegged Teku at 99.83%, while others indicated Lighthouse at 42.71%. The inconsistency in monitoring complicates the core developers’ ability to precisely target and support minority clients required for the necessary technical fragmentation.
Regulatory Pressure and MEV Infrastructure
The centralization risk has increased through the maximum extractable value (MEV) ecosystem, particularly the dependence on MEV-Boost infrastructure. Block confirmers utilize MEV-Boost to generate external relays and produce blocks that optimize staking profitability. By sourcing blocks from builders, these relays raise the total reward rate for the Block confirmer.
Censorship by Economic Incentive
The relay market is highly concentrated, with projects such as Flashbots dominating 81% of the total MEV-Boost. This concentration introduces a systemic compliance risk because relays often filter transactions to meet regulatory requirements, such as those imposed by the U.S. Office of Foreign Assets Control (OFAC). OFAC-compliant relays actively censor transactions involving sanctioned addresses, notably those related to the Tornado Cash smart contract.
The consequence of this voluntary adoption of profitable infrastructure is measurable censorship: 52% of blocks proposed through MEV-Boost relays have been observed enforcing OFAC sanctions in recent periods. Block confirmers are prioritizing the economic benefits (higher APR) afforded by these services over the network’s fundamental principle of transactional neutrality.
To counter this profit-driven centralization, ETH developers are advancing work on Proposer/Builder Separation (PBS). PBS viewks to divorce the Block confirmer’s role (proposer) from the transaction ordering party (builder), providing an architectural countermeasure to transactional censorship by democratizing the building process and introducing on-chain transparency.
Layer 2 Dependency
The rapid growth of (L2) scaling answers introduces an indirect centralization risk. L2 networks rely on the security of the ETH Layer 1 (L1). Consequently, L1 issues, such as congestion or censorship, can impact L2 performance.
Furthermore, L2 answers often launch with centralized components, primarily centralized sequencers, to ensure high speed and ahead functionality. The sequencer processes transactions and submits proofs to L1. While roadmaps exist for decentralizing these sequencers, reliance on a single, centralized operator in the interim allows for potential manipulation of transaction order, delays, or system exploitation. If the ecosystem where most user activity occurs is centralized at the sequencer level, the inherent decentralization of the L1 base layer is undermined for the end user.
Bottom Line
ETH is decentralized at its core protocol level; however, systemic economic incentives and regulatory pressures are manifesting in key infrastructure points—specifically the dominance of the liquid staking market and centralized MEV relays—which actively erode both its technical resilience and its censorship resistance. While developers have designed powerful architectural defenses such as PBS and protocol-enforced slashing, these mitigations must be implemented and adopted rapidly to reverse the current trend toward concentrated control and preserve the network’s long-term integrity and neutrality.
