Ethereum's shift to proof of stake in September 2022 created a pool of economic security: over 30 million ETH staked by validators who risk losing their deposit (slashing) if they behave maliciously. This security pool protects Ethereum, but it sits idle with respect to every other protocol.
New protocols that need decentralized validation face a bootstrapping problem. An oracle network, a data availability layer, or a cross-chain bridge needs validators, and those validators need economic stakes large enough to make attacks unprofitable. Building this security from scratch is expensive. Each new protocol must attract its own set of stakers, issue its own token for staking rewards, and hope that enough capital commits to make the system secure.
Restaking proposes a different model. Instead of building independent security, new protocols borrow it from Ethereum. Stakers who already have ETH committed to Ethereum's consensus opt in to additionally securing other services. The same capital backs multiple protocols simultaneously.
EigenLayer formalized this concept and built the infrastructure for it. This guide explains how restaking works, what EigenLayer introduced, and where the risks compound.
To understand restaking, start with what it extends.
Ethereum validators deposit 32 ETH into a staking contract. In return, they earn rewards for proposing and attesting to blocks (currently around 3% to 4% annualized). If a validator acts maliciously (double-signing, proposing conflicting blocks) or goes offline for extended periods, a portion of their 32 ETH is slashed.
This creates an economic security guarantee. Attacking Ethereum's consensus requires controlling enough staked ETH that the cost of being slashed exceeds the profit from the attack. With over 30 million ETH staked (roughly $100 billion at mid-2026 prices), that threshold is prohibitively high.
Liquid staking protocols like Lido (stETH) and Rocket Pool (rETH) added a layer on top. Users deposit ETH, receive a liquid token representing their stake, and can use that token in DeFi while still earning staking rewards. The underlying ETH remains staked with validators. For a detailed breakdown of how liquid staking tokens work and the depeg risks they carry, the mechanics are important context for understanding the additional risk layer that restaking introduces.
Restaking adds a second layer on top of staking (or liquid staking). The same ETH that secures Ethereum also secures additional protocols.
EigenLayer is a set of smart contracts on Ethereum that coordinate restaking. The system has three roles:
Restakers. Users who commit their staked ETH (or liquid staking tokens like stETH) to EigenLayer. Restakers deposit into EigenLayer's contracts and delegate their stake to an operator.
Operators. Entities that run validation software for actively validated services. An operator registers with EigenLayer, receives delegated stake from restakers, and opts into one or more AVSs. Operators are responsible for meeting each AVS's validation requirements and face slashing if they fail.
Actively validated services (AVSs). Protocols that use EigenLayer's restaked security. An AVS defines its own validation logic, reward structure, and slashing conditions. When an operator opts into an AVS, the restaked ETH backing that operator becomes subject to the AVS's slashing rules.
The flow:
EigenLayer's contracts enforce the delegation and slashing logic, but they do not define what constitutes a slashable offense. Each AVS writes its own slashing contract, which EigenLayer's DelegationManager calls when a slashing event is proven. This modularity is what allows any type of protocol to become an AVS, but it also means the security of each AVS's slashing logic varies independently.
AVSs are the demand side of the restaking marketplace. They are protocols that need decentralized validation but do not want to build their own validator set and token economy from scratch.
The first and largest AVS is EigenDA, a data availability layer built by EigenLayer's team. Rollups can post their transaction data to EigenDA instead of Ethereum's calldata or blobs, reducing costs while inheriting security from restaked ETH. By mid-2026, EigenDA was processing data for multiple L2 rollups, providing an alternative to Celestia and Ethereum's native blob space.
Other AVS categories include:
Oracle networks. A decentralized oracle can use restaked ETH as its security bond instead of requiring oracles to stake a separate token. If an oracle submits a false price, the restaked ETH backing it gets slashed. This provides stronger economic guarantees than a standalone oracle token with a small market capitalization.
Cross-chain bridges. Bridge validators can be backed by restaked ETH, creating an economic deterrent against fraudulent attestations far larger than what a standalone bridge token could provide. Given that bridge exploits have caused over $4 billion in losses, the appeal of Ethereum-grade security for bridge validation is significant.
Keeper networks. Protocols that require off-chain computation or automation (liquidation keepers, MEV relayers) can use restaked security to guarantee performance. An AVS slashing contract can penalize operators who fail to execute required actions within a time window.
Coprocessors. Off-chain computation services that produce verifiable results, such as ZK proof generation or AI inference verification, can use AVS slashing to enforce correct output. This category is expanding as more protocols look to verify off-chain computation without running it on-chain.
By mid-2026, over 20 AVSs had launched on EigenLayer, with EigenDA processing the highest volume. EigenLayer's expansion to accept any ERC-20 token as a restakable asset broadened the potential collateral base beyond ETH and its liquid staking derivatives.
Just as liquid staking created tradable representations of staked ETH (stETH, rETH), liquid restaking protocols create tradable tokens representing restaked positions.
The major liquid restaking protocols:
Ether.fi (eETH). The largest liquid restaking protocol by TVL. Users deposit ETH, Ether.fi stakes it and restakes it through EigenLayer, and users receive eETH that they can use across DeFi. Ether.fi outpaced competitors in the liquid staking sector by offering a streamlined one-step deposit flow and integrating with major DeFi protocols for composability.
Renzo (ezETH). Abstracts the EigenLayer delegation process. Users deposit ETH or stETH, Renzo handles operator selection and AVS opt-in, and users receive ezETH. Renzo differentiates by offering diversified AVS exposure: the protocol spreads delegated stake across multiple operators and AVSs to reduce concentration risk.
Puffer (pufETH). Focuses on solo validator participation and anti-slashing technology alongside liquid restaking. Puffer's approach includes secure-signer technology that aims to prevent validators from producing slashable messages, even if their keys are compromised.
Kelp (rsETH). Aggregates restaked positions across operators and AVSs into a single liquid token. Kelp aims to provide diversified restaking exposure similar to an index fund approach.
LRTs add convenience but also add another layer of smart contract risk. The stack becomes: ETH -> staked ETH -> liquid staking token -> restaked on EigenLayer -> liquid restaking token. Each layer introduces its own contract, its own governance, and its own potential failure mode. A bug or exploit at any layer can cascade downward.
You might also like: What is proof of stake? How validators replaced miners
Restaking's value proposition depends on simple economics.
Suppose a new oracle network needs $100 million in economic security to make attacks unprofitable. Without restaking, it must convince stakers to buy and lock $100 million worth of its native token. The token needs price stability, liquidity, and market confidence, none of which a new project has on day one.
With restaking, the oracle network becomes an AVS on EigenLayer. It borrows security from ETH already staked, a liquid asset with deep markets and established value. The oracle does not issue a staking token. It pays ETH-denominated rewards to operators, and the $100 million in restaked ETH backing those operators provides the security.
The cost to the AVS is the reward it must pay operators (and by extension restakers) to opt in. This is typically denominated in the AVS's own token or in ETH. The cost is lower than bootstrapping a standalone staking economy because restakers already earn base staking yield. The AVS only needs to offer enough marginal reward to justify the additional slashing risk.
For restakers, the appeal is yield stacking. A position might earn:
Aggregate yields of 5% to 7% on ETH drew significant capital into restaking during 2024 and 2025. At its peak, EigenLayer held over $15 billion in restaked assets, making it one of the largest DeFi protocols by TVL.
However, yield stacking is not free money. Each additional percentage point of yield comes with a corresponding increase in risk exposure. The higher the aggregate yield, the more slashing vectors the position is exposed to.
The compounding of yield comes with compounding of risk. Restaked ETH is subject to slashing from multiple sources simultaneously.
Ethereum consensus slashing. If the underlying validator double-signs or commits an attributable fault, the base stake is slashed under Ethereum's rules. This risk exists with or without restaking.
AVS slashing. Each AVS the operator opts into introduces its own slashing conditions. An operator running three AVSs faces three independent sets of slashing rules. A bug in any single AVS's slashing contract could trigger an incorrect slash.
Correlated slashing. If an operator runs multiple AVSs and a single infrastructure failure (a data center outage, a key compromise) causes violations across all of them, the same stake can be slashed multiple times. EigenLayer's contracts permit proportional slashing, meaning the total slash can exceed what would occur from any single AVS.
Smart contract risk in slashing contracts. AVS slashing logic is defined in smart contracts written by the AVS team. A bug in the slashing contract could slash honest operators. Unlike Ethereum's consensus slashing, which has been battle-tested since 2020, AVS slashing contracts are new and less audited.
LRT compounding risk. Users holding liquid restaking tokens face all the above risks plus the smart contract risk of the LRT protocol itself, and the risk that the LRT depegs from its underlying value during a slashing event or a liquidity crisis.
Systemic risk. If a large-scale slashing event hits a major operator, the resulting sell pressure on LRTs could trigger cascading liquidations in DeFi protocols that accept LRTs as collateral. A restaking-linked liquidation cascade has not occurred yet, but the structural possibility exists as more DeFi protocols integrate LRTs as collateral types.
Restaking is no longer an EigenLayer monopoly.
Symbiotic launched in 2024 as a permissionless restaking protocol. Unlike EigenLayer, which initially only accepted ETH and liquid staking tokens, Symbiotic accepts any ERC-20 token as collateral. This allows protocols to restake their own governance tokens or stablecoins. Symbiotic's architecture is also more modular: slashing conditions, reward distribution, and operator management are separated into distinct contracts that each AVS can customize independently.
Karak introduced the concept of restaking across multiple chains, with support for restaking on Arbitrum, Mantle, and other L2s in addition to Ethereum mainnet. Karak's multi-chain approach appeals to AVSs that want security from assets on chains other than Ethereum, and to restakers who want to avoid bridging to Ethereum mainnet.
Babylon applies the restaking concept to Bitcoin. BTC holders lock their Bitcoin in a time-locked script and use it to secure proof-of-stake chains. The Bitcoin never leaves the Bitcoin blockchain (no wrapping, no bridging), but it is subject to slashing via a cryptographic penalty mechanism called extractable one-time signatures. If a staker signs conflicting messages, the EOTS scheme reveals their private key, allowing anyone to claim the locked Bitcoin as a penalty.
The emergence of competitors suggests that restaking is becoming a category, not a single product. The long-term question is whether security fragmentation across competing restaking layers weakens the shared security model that makes restaking valuable in the first place. If the same capital is split across EigenLayer, Symbiotic, and Karak, the security each provides is proportionally reduced.
Not all EigenLayer operators carry the same risk profile. The choice of operator determines which AVSs your stake is exposed to, the quality of the infrastructure running those AVSs, and the operational maturity of the team managing the node.
Professional operators (Figment, P2P, Kiln, and similar institutional staking providers) typically run redundant infrastructure across multiple data centers, maintain dedicated security teams, and limit the number of AVSs they opt into. Solo operators or smaller teams may offer higher yields by opting into more AVSs, but they also concentrate risk in fewer hands and less resilient infrastructure.
The operator's track record is the most reliable signal. EigenLayer's delegation dashboard shows historical uptime, slashing events (if any), and the list of active AVS commitments. An operator with 99.9% uptime across 12 months of operation and a conservative AVS selection provides a meaningfully different risk profile than a new operator running aggressive multi-AVS strategies.
Delegation is not permanent. Restakers can re-delegate to a different operator, though the process involves a withdrawal delay. If an operator begins opting into AVSs with unclear slashing conditions or questionable audit histories, re-delegation is the primary risk management tool available to restakers.
This guide explains restaking mechanics and risks. It does not cover:
Understand operator risk. When you delegate to an operator, you inherit their slashing exposure. Review which AVSs the operator has opted into, their uptime history, and their infrastructure setup. An operator running 15 AVSs on a single server in a single data center is a concentrated risk.
Review AVS slashing conditions. Before your operator opts into a new AVS, understand what triggers a slash. Some AVS slashing conditions are straightforward (fail to submit data within a window). Others are complex or depend on dispute resolution mechanisms that have not been tested under stress.
Assess LRT risks separately. If you hold a liquid restaking token, you carry the restaking risk plus the LRT protocol's smart contract risk. Check audit reports for both the LRT protocol and the underlying restaking contracts. Consider the LRT's redemption mechanism: some LRTs allow instant redemption, while others queue withdrawals.
Monitor your position. Restaking is not a deposit-and-forget strategy. New AVSs, operator changes, and slashing events can alter your risk profile. Protocols like EigenLayer provide dashboards showing operator performance and AVS status. Set up notifications for operator changes if the protocol supports them.
Consider the withdrawal queue. Restaked positions may have longer withdrawal periods than simple staking. EigenLayer enforces a withdrawal delay (currently 7 days), and during high-demand periods the queue can extend. Do not restake funds you may need to access quickly. Factor withdrawal timing into your liquidity planning.
Read more: What are liquid staking tokens? stETH and the depeg risk, explained
Restaking means using ETH that is already staked on Ethereum to simultaneously secure other protocols. The same deposit earns staking rewards from Ethereum and additional rewards from the other protocols it helps secure, in exchange for accepting additional slashing risk.
An actively validated service (AVS) is a protocol that uses restaked ETH from EigenLayer for its security. Examples include data availability layers, oracle networks, bridges, and keeper networks. Each AVS defines its own validation requirements and slashing conditions.
Liquid staking (Lido, Rocket Pool) creates a tradable token representing staked ETH. The ETH secures only Ethereum's consensus. Restaking takes that staked ETH and commits it to securing additional protocols beyond Ethereum. Liquid restaking combines both: it creates a tradable token representing a restaked position.
Yes. Restaked ETH is subject to slashing from Ethereum's consensus rules and from every AVS the operator has opted into. If the operator behaves maliciously or suffers a fault that triggers AVS slashing conditions, a portion of the restaked ETH can be permanently destroyed.
Returns vary by operator and AVS. Base Ethereum staking yields approximately 3% to 4%. AVS rewards can add 1% to 3% or more, depending on the service. Total yields of 5% to 7% were common during 2024 and 2025, though these fluctuate with market conditions and AVS demand.
Restaking introduces additional risk layers beyond standard staking. Each AVS adds a new slashing vector, and the slashing contracts are newer and less battle-tested than Ethereum's consensus penalties. Operator selection, AVS due diligence, and smart contract audit quality all affect the safety of a restaking position.
A liquid restaking token (LRT) is a tradable token representing a restaked position. Protocols like Ether.fi (eETH), Renzo (ezETH), and Puffer (pufETH) issue LRTs that let users maintain DeFi composability while their ETH is restaked. LRTs carry the underlying restaking risk plus the LRT protocol's own smart contract risk.
Yes, through Babylon Protocol. BTC holders lock Bitcoin in a time-locked script on the Bitcoin blockchain (no wrapping or bridging required) and use it to secure proof-of-stake chains. Slashing is enforced through a cryptographic mechanism that extracts the staker's private key if they sign conflicting messages.
*Disclaimer: This article is for informational purposes only and does not constitute financial, investment, or legal advice. Cryptocurrency involves significant risk, and you should conduct your own research before making any decisions. Information is accurate as of August 2026.*
This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.















Banco Santander's Form 13F-HR, filed July 29, 2026, reports 129,615 shares of BlackRock's iShares Bitcoin Trust worth $4.31M as of June 30 — its first-ever disclosed spot-bitcoin-ETF position, alongside a first ether-ETF holding. What the filing shows is a quarter-end snapshot of ETF shares; what it does not show is motive, or bitcoin on the bank's balance sheet.




Agave v4.2 targets mainnet feature activations on 17 August 2026, but the slot-time change is only step one of a five-step reduction — 400ms to 350ms, not the 200ms end state. Alpenglow's complete code ships with this release; its activation is set for v4.3.











Japan and the United States bought yen together on 31 July 2026, their first coordinated yen-buying operation since June 1998, and the currency has since firmed from a 40-year low of 163.99 into the 155-156 area. This explainer sets out the three channels linking the yen to bitcoin: carry-trade unwind risk, yen-denominated repricing, and the dollar-weakness correlation that points the other way.