Restaking is built on a seductive premise: ETH already staked to secure Ethereum can simultaneously secure other protocols, generating additional yield without requiring new capital. EigenLayer pioneered this model on Ethereum, and by mid-2025 it had attracted tens of billions in restaked ETH. The pitch is efficient capital deployment. The reality is that you’re stacking slashing conditions on top of each other, and most participants haven’t modeled what that looks like when something breaks.
What Restaking Actually Does
When a validator restakes, they opt into one or more Actively Validated Services (AVSs) - external protocols that need economic security but don’t have their own validator set. The restaked ETH acts as collateral: if the validator misbehaves on an AVS, some of that ETH can be slashed. The validator earns fees from the AVS in exchange.
This sounds like a clean extension of proof-of-stake logic. The problem is that slashing risk compounds non-linearly. A validator opting into five AVSs isn’t taking five separate, independent bets - they’re creating a situation where a bug, an exploit, or even ambiguous slashing criteria in any one AVS can trigger losses that affect the underlying Ethereum stake. The security guarantee Ethereum offers is diluted every time that stake is pledged elsewhere.
The Correlation Problem
Ethereum’s economic security works because slashing is rare, predictable, and governed by clear rules. Restaking breaks that predictability. AVSs vary enormously in their technical maturity, governance quality, and slashing logic. Some are audited; many are not. A validator who has restaked across several AVSs is exposed to the weakest link in that chain - and unlike diversified financial portfolios, correlated failure in restaking can cascade rather than cancel out.
Operators and delegation protocols like EigenLayer’s operator marketplace add another layer. Retail stakers who delegate to operators often don’t know which AVSs their restaked ETH is committed to, or what the slashing parameters look like.
The Yield Obscures the Mechanism
The additional yield restaking generates is real. So is the fact that most depositors are not evaluating AVS slashing contracts before opting in.
What restaking has done, at scale, is introduce a new category of systemic risk into Ethereum’s validator set - one that isn’t visible in block explorers or standard staking dashboards. Ethereum’s base-layer security is only as robust as the assumptions validators make about the obligations they’ve layered on top of it. How fragile those assumptions are won’t be known until the first major AVS failure stress-tests them for real.