Restaking after slashing went live

Staked assets secure their base network and earn its issuance. Restaking commits the same stake to secure additional services, such as data availability layers, oracle networks, bridges and sequencers, in exchange for additional reward. The stake becomes exposed to the slashing conditions of every service it secures, on top of the base network’s. That layering is both the source of the extra yield and the reason it exists.

The reality check

For most of the sector’s history, slashing on the largest restaking protocol was not enabled. Delegators earned rewards for a risk that could not, in practice, be realized. On 17 April 2026 EigenLayer activated production slashing on mainnet. Industry coverage put the restaked value exposed to real cryptoeconomic loss for the first time in the mid-teens of billions of dollars. From that date, an operator fault on any service a delegation secures can cost the delegator stake.

Isolation

The design question that matters most is whether a fault on one service can burn stake committed to another. The newer allocation model on EigenLayer isolates stake per service, so that a buggy slashing condition on one cannot reach unrelated delegations through a shared operator. Older delegations may not have that protection. Protofire’s 2026 comparison of restaking protocols and Blockdaemon’s institutional review both treat this isolation, and the equivalent design choices on Symbiotic, as the first question an institution should ask before delegating.

• Is slashing live on this protocol, and what are the exact conditions per service?
• Is stake allocated to each service isolated, or shared across everything the operator secures?
• Who is the operator, what infrastructure does it run, and what is its public record?
• What does each service pay relative to what it can slash?
• What are the unbonding periods, and what discount do the liquid tokens trade at in stress?

A smaller, more serious market

Total value in restaking has fallen substantially from its peak, and several accounts attribute part of the decline to the arrival of real risk. Institutional operators have stayed through the drawdown, which VaaSBlock and others read as a signal of commitment. A smaller market with real slashing is, for an allocator, a more honest one: the yield now compensates for a risk that can actually occur, and the diligence has something concrete to test.

What changes for an allocator

Restaking should be sized as a position that can lose principal, not as a yield enhancement on a staked asset. The compounding case, base slashing plus every service’s maximum penalty, sets the exposure limit. Delegation goes to operators with institutional infrastructure. Services are reviewed individually, and any that can slash across an entire delegation are declined regardless of reward. Liquid restaking tokens used elsewhere as collateral are haircut to their worst observed discount, because the exit from a restaked position runs through an unbonding queue that does not open faster in a crisis.