EigenLayer and Restaking: How Ethereum Is Securing the Next Internet Layer

Ethereum staking earns approximately 3–4% annual yield for securing the network. What if the same staked ETH could simultaneously earn yield from securing five other protocols — a data availability layer, an oracle network, a bridge, a rollup sequencer? This is EigenLayer's proposition: "restaking," the ability to reuse staked ETH as security collateral for additional protocols, earning supplemental yield on capital already at work. Since launching on mainnet in mid-2023, EigenLayer has accumulated $15B+ in restaked assets and spawned a $10B+ liquid restaking ecosystem. It's one of the most significant DeFi innovations since liquid staking itself — and one of the most actively debated from a systemic risk perspective. Here's everything you need to know.

Why Restaking Exists: The Security Bootstrapping Problem

Building a new decentralised protocol that requires economic security faces a bootstrap problem. You need validators with significant capital at stake to make attacks economically irrational — but those validators require rewards, which requires the protocol to have revenue, which requires users, which requires the protocol to be functional and trusted, which requires security. New protocols typically start with low security (small validator sets, limited staked capital) and grow security as they grow, spending early years vulnerable to attacks. This is why many bridges have been exploited: insufficient economic security against well-capitalised attackers. EigenLayer breaks this cycle by allowing new protocols to rent Ethereum's validator security from day one — accessing $100B+ in economic security without waiting years to bootstrap their own validator communities.

The mechanism: operators (Ethereum validators or delegated staking services) opt in to Actively Validated Services (AVSs) on EigenLayer. By opting in, they agree to run the AVS's software and follow its rules, with additional slashing conditions for misbehaviour. AVSs pay operators for their security service — either through their native tokens or fee revenue. Restakers who delegate to operators also share in this additional yield proportional to their delegation. The AVS gets security; the restaker and operator get additional yield; Ethereum's validator set gains additional productive use of their capital. The theory is elegant. The implementation is the interesting part.

How to Restake: Native vs LST Restaking

There are two ways to restake into EigenLayer. Native restaking requires running your own Ethereum validator and pointing the validator's withdrawal credentials to EigenLayer smart contracts. This gives EigenLayer authority over a portion of your withdrawal queue, enabling slashing enforcement. Native restaking provides the most direct participation but requires a minimum 32 ETH validator setup and technical node operation. LST restaking is simpler: deposit stETH, rETH, cbETH, or other liquid staking tokens directly into EigenLayer's strategy contracts. The LST's value is now subject to EigenLayer's slashing conditions in addition to Ethereum validator slashing. No node operation required — you're delegating to operators who run the actual AVS software. For most retail participants, LST restaking (directly or through a liquid restaking token protocol) is the accessible path.

Direct EigenLayer restaking requires managing which operators to delegate to and which AVSs to opt into — decisions with different risk/reward profiles for each AVS. Most users prefer the abstraction layer provided by Liquid Restaking Token protocols.

Liquid Restaking Tokens: The Simplified Interface

Liquid Restaking Tokens (LRTs) abstract EigenLayer complexity into a single deposit experience. You deposit ETH or stETH, receive an LRT that represents your restaked position, and the protocol manages operator selection, AVS diversification, and reward compounding on your behalf. The LRT appreciates over time as both LST staking yield and EigenLayer AVS rewards accumulate in the exchange rate. Major LRT protocols in 2026:

ether.fi (eETH/weETH): The largest LRT by TVL ($5B+). ether.fi operates its own native Ethereum validators (not relying on third-party operators), providing direct security over the validator infrastructure. The ETHFI governance token and significant airdrop distributions have made ether.fi the most capitalised LRT protocol. Unusually, ether.fi also offers a debit card that lets users spend DeFi yield in real-world purchases — an ambitious product expansion beyond pure restaking. Puffer Finance (pufETH): Uses a novel "secure signer" architecture that allows node operators to run Ethereum validators with only 2 ETH collateral (versus 32 ETH normally) by using anti-slashing hardware guards. This democratises native restaking to smaller operators. Renzo (ezETH): Focuses on multi-AVS diversification, automatically spreading restaked capital across multiple AVSs to optimise yield while limiting concentration in any single AVS's risk. The RNDO governance token with retroactive airdrop distribution generated significant initial TVL. Kelp DAO (rsETH): Multi-asset LRT accepting stETH, ETHx, and sfrxETH as deposits, appealing to holders of various LST types who want restaking exposure without converting to a single LST first.

EigenDA: The Flagship AVS and Revenue Model

EigenDA is EigenLayer's own data availability service — the most prominent AVS and the clearest demonstration of what restaking enables. Rollups (Ethereum L2s) need to make their transaction data available so anyone can reconstruct the rollup state and verify state transitions. Posting data to Ethereum mainnet as blobs costs gas even after EIP-4844's improvements. EigenDA provides data availability at 10–100x lower cost than Ethereum blobs by distributing data across its own node network (secured by restaked ETH) rather than all Ethereum full nodes. The security guarantee: the restaked ETH behind EigenDA operators means data withholding attacks require corrupting validators with $100B+ in total economic stake — the same threshold as attacking Ethereum itself.

EigenDA's clients in 2026 include Celo's L2 migration, Mantle, and several emerging rollup projects. As rollup data volumes increase with EVM adoption, EigenDA's fee revenue should grow proportionally — creating genuine demand-side income for restakers rather than purely inflationary token rewards. The progression from "theoretical yield from future AVS fees" to "actual fee revenue from deployed AVSs with production traffic" is the maturation milestone that distinguishes sustainable restaking economics from speculative allocation.

The Risk Profile: What You're Taking On

Restaking's additional yield comes with additional risk — understanding both components is essential before committing capital. The primary risks:

AVS slashing risk: If an operator you've delegated to misbehaves on an AVS (signs conflicting attestations, goes offline during a critical window, executes a prohibited action), the AVS's slashing conditions can reduce your restaked balance. EigenLayer's "slashing manager" contract executes these reductions. In practice, slashing events have been limited since launch — most AVSs have conservative early slashing parameters — but the risk scales with the number and aggressiveness of AVSs you're exposed to through your LRT's operator portfolio. Smart contract risk: EigenLayer and each LRT protocol represent additional smart contract layers above the base LST risk. Multiple audits have been conducted, but the attack surface is substantially larger than a simple LST deposit. LRT depeg risk: LRTs can depeg from ETH just as stETH depegged in June 2022 — if a large slashing event occurs, or if market fear of slashing causes panic selling of LRT tokens, the LRT/ETH exchange rate can break significantly below 1:1. Leveraged positions using LRT collateral face liquidation risk from both ETH price drops and LRT depegs simultaneously.

Systemic concern (Vitalik's argument): Ethereum's social consensus — the ability to coordinate a community response to catastrophic events — should not be used to backstop AVS failures. If a large AVS is slashed and the slashing is contested, there's a risk that AVS participants pressure the Ethereum community to fork the chain to reverse the slashing. This would pollute Ethereum's neutrality as a base layer. EigenLayer has implemented governance controls (veto committees that can pause slashing in case of bugs) but the fundamental tension between AVS security relying on Ethereum's bedrock security and Ethereum's core protocol remaining free from AVS-specific disputes hasn't been fully resolved architecturally. Approach restaking as a 5–15% satellite allocation with explicit slashing risk budget, not as a replacement for lower-risk staking positions.

Practical Yield Math in 2026

A realistic restaking yield stack in 2026: base ETH staking yield (3.5%) + EigenLayer AVS rewards (1–3% depending on AVS portfolio and EIGEN token incentives) + LRT protocol incentives (0.5–2% in early growth phase) = total 5–8.5% APY. This compares favourably to native ETH staking (3.5%) and is competitive with stablecoin yields while maintaining ETH price exposure. The yield premium over native staking (1.5–5%) is the compensation for accepting additional slashing risk and smart contract risk. Whether this risk premium is adequate depends on your assessment of AVS slashing probability across your LRT's operator portfolio — a judgment call that requires researching which AVSs your chosen LRT is exposed to and their individual slashing risk parameters. For most long-term ETH holders, allocating 20–40% of staking exposure to an LRT while keeping the remainder in a well-established LST (stETH, rETH) provides a reasonable balance of additional yield and risk diversification.