How Restaking Works: EigenLayer’s Impact on Ethereum Security

If you’ve been watching Ethereum’s evolution, you’ve noticed that staking has become central to the network’s security model. What you may not have followed as closely is how a relatively new mechanism called restaking has changed that security architecture — and why EigenLayer’s approach has sparked both excitement and controversy across the ecosystem.

Restaking isn’t just another DeFi trend. It represents a shift in how Ethereum thinks about cryptoeconomic security, extending the guarantee that validators provide to the base layer outward into application-specific services that never had access to that level of economic commitment before. The implications reach far beyond yield farming.

This article breaks down exactly how restaking functions, why EigenLayer’s implementation matters, what security guarantees actually change, and where the risks lie. I’ll be direct about what’s genuinely innovative here and what deserves healthy skepticism.

What is Restaking in Ethereum?

Restaking allows ETH stakers to validate additional protocols beyond Ethereum’s native consensus layer without committing additional capital. Instead of running a separate validator for each network, stakers reuse their existing staked ETH to secure multiple services simultaneously. The economic security of the base layer — billions of dollars in staked value — gets applied across an expanding set of applications.

The mechanism gained prominence through EigenLayer, a protocol that formalized restaking as a programmable primitive. When you restake through EigenLayer, your ETH secures what are called Actively Validated Services (AVSs) — decentralized systems that need their own consensus or validation mechanisms. These could be data availability layers, cross-chain bridges, oracle networks, or entirely new blockchain architectures.

Here’s what makes this significant: before restaking, each new protocol had to bootstrap its own security from scratch. Building a validator set required convincing people to stake capital specifically for that protocol, creating a cold-start problem that many ambitious projects never solved. Restaking effectively lets Ethereum’s stakers port their economic weight anywhere it’s needed.

The total value locked in restaking protocols has grown substantially since EigenLayer’s mainnet launch in February 2024. According to DefiLlama, EigenLayer reached approximately $15 billion in TVL by late 2024, making it one of the fastest-growing DeFi primitives in Ethereum’s history.

How Traditional Staking Works

To understand what restaking changes, you need to understand what traditional staking actually does — and where its limitations lie.

Ethereum’s Proof of Stake consensus requires validators to deposit 32 ETH to participate in block production. These validators propose and attest to blocks, and their staked ETH serves as collateral that can be slashed (partially or entirely destroyed) if they behave dishonestly. This creates cryptoeconomic security: validators have skin in the game, and attacking the network becomes astronomically expensive relative to any potential profit.

The network currently has over 900,000 validators with approximately 29 million ETH staked as of early 2025, representing tens of billions of dollars in economic security. This makes Ethereum one of the most secure blockchains by pure economic commitment.

The limitation isn’t security — it’s flexibility. That massive economic guarantee only protects Ethereum’s base layer. If you wanted to build a bridge between Ethereum and another chain, that bridge needed its own validator set, its own security model, and its own economic guarantees. The security didn’t stack; it siloed.

This created a fragmented security landscape where protocols constantly reinvented the wheel, and where smaller projects struggled to achieve meaningful economic security. A bridge with $50 million in TVL might have only $5 million in validator stakes protecting it — an attractive target for attackers.

What is EigenLayer and How Does It Work?

EigenLayer is a restaking protocol that creates a marketplace for cryptoeconomic security. It allows stakers to opt into securing third-party protocols by staking their ETH (or liquid staking tokens) through the EigenLayer smart contracts. In return, they earn additional yields from these protocols.

The mechanism works through a few interconnected components:

Restaking: ETH holders can restake their stETH (from Lido), rETH (from Rocket Pool), or native staked ETH directly into EigenLayer’s contracts. This reuses their existing stake to secure additional services.

Actively Validated Services (AVSs): These are the protocols that benefit from restaking. An AVS defines its own validation requirements and slashable offenses, then taps into EigenLayer’s pool of restaked ETH for security.

EigenPods: These are validator-specific smart contracts that enable restaking for those running native Ethereum validators. EigenPods track validator balances and enable partial restaking configurations.

Slasher: EigenLayer implements its own slashing mechanism that can penalize restakers who violate AVS rules, creating a direct link between restaking participation and protocol-specific penalties.

What makes EigenLayer architecturally interesting is its approach to “free delegation.” Restakers don’t need to actively operate infrastructure for AVSs — they simply commit their stake and accept the slashing conditions. Operators then do the actual validation work, splitting rewards with the restakers whose capital they use.

This creates a separation: capital providers earn yields for bearing slashing risk, while operators earn fees for running infrastructure. The protocol facilitates the match without requiring each side to handle the other’s responsibilities.

How EigenLayer Changes Ethereum Security

The security implications are both substantial and nuanced. Here’s where the conversation gets interesting, because there’s genuine disagreement among security researchers about whether restaking strengthens or weakens Ethereum’s overall security model.

The Case for Strengthening Security

Restaking dramatically increases the economic cost of attacking interconnected systems. When a bridge or oracle network taps into restaked ETH, attackers now face the combined economic security of both the base protocol and EigenLayer’s restaked pool. This is meaningfully different from protocols with small, dedicated validator sets.

EigenLayer also enables what Vitalik Buterin has called “vertical integration of security” — the ability to pool security resources rather than fragmenting them. Instead of ten different bridges each trying to secure $100 million with inadequate validator sets, they can all draw from the same multi-billion-dollar security pool.

Projects like EigenDA (a data availability layer) and various rollups have already integrated with EigenLayer to benefit from this shared security model. The economics are appealing: paying a portion of your revenue to tap into Ethereum’s security is often cheaper than building your own validator economy.

The Genuine Tension

Here’s the point that many articles gloss over: restaking can actually increase systemic risk even as it improves individual protocol security.

When the same ETH secures multiple protocols, a catastrophic failure in one could cascade. If a major AVS gets exploited and triggers mass slashing, those same restakers might be providing security to other AVSs simultaneously. The interconnectedness that makes restaking powerful also creates correlation risk.

Ethereum’s security model relies on the assumption that validators have a long-term time preference — they’re incentivized to maintain the network because they hold staked ETH that would be destroyed if they attacked the system. Restaking adds shorter-term yield incentives on top of that, and researchers at ChainSecurity have raised concerns about whether this shifts validator incentives in ways that could matter during crisis scenarios.

The honest assessment is that we don’t fully know how this plays out. Restaking hasn’t been stress-tested by a major market crash or a sophisticated attack scenario. The theoretical models suggest manageable risk; practice may differ.

Risks and Criticisms of Restaking

Any honest treatment of restaking needs to address the substantial criticisms that have emerged from within the Ethereum community.

Smart Contract Risk: Restaking introduces additional smart contract exposure. You’re trusting not just Ethereum’s core contracts but also EigenLayer’s implementation and whatever AVSs you indirectly support. Multiple layers of smart contracts mean multiple potential failure points.

Centralization Concerns: EigenLayer has a significant operator concentration issue. A relatively small number of operators control most of the restaked ETH, and this has drawn criticism from researchers concerned about permissioned sets and the potential for coordination failures. The team has announced plans to address this, but it’s a current limitation.

Liquidity Risks: Liquid restaking tokens (LRTs) introduce complex liquidity dynamics. When everyone rushes to exit LRT positions during market stress, the underlying ETH might be locked in EigenLayer contracts, creating potential redemption delays or discounts.

AVS Selection Risk: Not all AVSs will be equally secure. Restakers need to evaluate which services they’re implicitly backing, and there’s concern that yield-seeking behavior might lead capital toward riskier protocols simply because they offer higher rewards.

The most credible criticism is that restaking optimizes for economic efficiency at the cost of redundancy. If multiple critical infrastructure pieces draw from the same security pool, a single point of failure could take down more than one system simultaneously. This is the opposite of defense in depth.

The Liquid Restaking Token (LRT) Ecosystem

The restaking narrative has spawned an entire ecosystem of liquid restaking protocols. These protocols let users restake through them and receive a liquid token in return — essentially combining staking derivatives with restaking functionality.

Renzo (ezETH): One of the largest LRT protocols, Renzo has accumulated billions in TVL by aggregating restaking deposits and offering users an easy entry point to EigenLayer yields.

Swell (swETH): Another significant LRT that has grown rapidly, competing on yield optimization and user experience.

Kelp (rsETH): Offers a more permissioned approach to restaking, with a focus on security and risk management.

The LRT ecosystem has grown contentious because these tokens trade at various discounts to their underlying ETH value, creating opportunities and risks that are still being understood. During periods of market stress in late 2024, some LRTs experienced significant depegs, raising questions about the liquidity infrastructure supporting this market.

Conclusion: The Future of Restaking on Ethereum

Restaking represents one of the most significant architectural evolutions in Ethereum’s history, but we’re still in the early stages. The technology offers real solutions to genuine problems — security fragmentation, validator coordination costs, and capital inefficiency — while introducing trade-offs that the ecosystem is still learning to manage.

What I find most interesting is the philosophical shift underneath. Ethereum has traditionally approached security as something monolithic — the entire network secures itself. Restaking introduces a more modular view, where security becomes a composable resource that can be allocated where it’s most needed. Whether this leads to a more resilient or more fragile ecosystem remains to be seen.

If you’re considering participating in restaking, the practical takeaway is straightforward: understand what you’re actually backing. The yields are attractive, but they’re compensation for taking on smart contract risk, slashing risk, and systemic correlation risk. These aren’t negligible considerations.

The next year will reveal whether restaking achieves the security compounding its proponents envision or whether the risks materialize in ways that force architectural changes. For now, the experiment is underway, and it’s one worth watching closely.