EigenLayer crossed 5 million ETH in restaking deposits in early September 2026, cementing its position as the dominant shared-security protocol on Ethereum with approximately 93.9% market share of the restaking sector. The protocol, rebranded under the EigenCloud umbrella in mid-2025, now secures...
"We should preserve the chain's minimalism, support uses of restaking that do not look like slippery slopes to extending the role of Ethereum consensus." — Vitalik Buterin, Ethereum Co-Founder
EigenLayer crossed 5 million ETH in restaking deposits in early September 2026, cementing its position as the dominant shared-security protocol on Ethereum with approximately 93.9% market share of the restaking sector. The protocol, rebranded under the EigenCloud umbrella in mid-2025, now secures 18 actively validated services (AVSs) with roughly $19.7 billion in total value locked.
The growth trajectory is undeniable. So are the risks. The April 2026 Kelp DAO bridge exploit — $292 million drained through compromised RPC nodes, not a smart contract bug — demonstrated how liquid restaking tokens (LRTs) create contagion pathways that did not exist 18 months ago. An IEEE study published in 2026 found that many AVSs are secured by highly centralized validator sets, undermining the decentralized security thesis that underpins the entire restaking model. Meanwhile, the EIGEN token trades at $0.21, down 96% from its all-time high of $5.65, raising questions about the gap between protocol adoption and token value capture.
EigenLayer's restaking deposits crossed 5 million ETH in the first week of September 2026, according to data reported by Bitcoinist and CoinDesk. The figure includes both native ETH and liquid staking token (LST) deposits across the protocol's operator set.
To contextualize: 5 million ETH represents approximately 4.2% of Ethereum's total supply and roughly 12.3% of all staked ETH. According to protocol data, EigenLayer commands 93.9% of all capital in the restaking sector, with $19.7 billion in total value locked spread across 4.36 million ETH in operator-delegated positions. The remaining 6.1% is distributed across smaller protocols including Symbiotic and Karak.
The deposit growth occurred despite — or perhaps because of — declining base staking yields. With 34% of all ETH now staked, according to Ethereum staking statistics, base validator yields have compressed to approximately 2.8-3.2% APY. Restaking offers supplemental yield from AVS fees, pushing combined returns to the 8-12% APY range for operators participating in multiple services. The yield premium explains the capital flow, though it comes with additional slashing exposure.
On September 1, 2026, 39.5 million EIGEN tokens (4.49% of circulating supply, valued at approximately $7.66 million) were unlocked from vesting schedules allocated to early backers and staff. Bitfinex delisted the EIGEN token effective August 31, 2026.
In mid-2025, Eigen Labs rebranded from EigenLayer to EigenCloud, integrating EigenDA (data availability), EigenVerify (verification services), and EigenCompute (off-chain computation) under a unified product stack. The rebrand signals a strategic pivot: positioning restaked ETH not merely as collateral for third-party protocols, but as the security substrate for a full "verifiable cloud" infrastructure.
The company describes EigenCloud as "a verifiable cloud for the agentic era," according to its official blog — infrastructure for autonomous AI agents that require cryptographic proof of computation. The positioning draws a direct comparison to AWS, but with verifiability baked in at the protocol level.
As of September 2026, more than 20 AVSs have launched on EigenCloud. EigenDA handles the highest throughput, serving as a data availability layer for rollups seeking lower costs than Ethereum's native blob space. Production slashing went live on mainnet on April 17, 2026, according to reporting from BlockEden. For the first time, the roughly $15 billion in restaked ETH faces enforceable cryptoeconomic penalties for operator misbehavior.
The activation of slashing is significant because it converts restaking from a yield-extraction mechanism into an actual security commitment. Before April 2026, operators could collect AVS fees without facing real financial penalties for poor performance. That asymmetry has been partially corrected, though EigenLayer's proposed operator safety scores and AVS risk rating systems remain under development as of September 2026.
The liquid restaking token (LRT) sector — protocols that issue tradable tokens representing restaked positions — has consolidated around a handful of providers. According to DefiLlama data, the sector holds approximately $4.6 billion across 28 tracked protocols.
Market share breakdown as of mid-2026:
| Protocol | Token | TVL | Market Share | |----------|-------|-----|-------------| | Ether.fi | eETH | ~$2.8B | ~61% | | Kelp DAO | rsETH | ~$874M | ~19% | | Renzo | ezETH | ~$89M | ~2% | | Puffer Finance | pufETH | ~$43M | ~1% | | Others (24 protocols) | Various | ~$789M | ~17% |
Ether.fi dominates with roughly 61% of LRT TVL. The protocol abstracts EigenLayer complexity: users deposit ETH, receive eETH, and earn both base Ethereum staking yields and restaking rewards without managing operator selection. This simplification drives adoption but concentrates risk. If Ether.fi's operator selection algorithm misdirects capital toward an AVS that suffers a slashing event, losses propagate to all eETH holders simultaneously.
Renzo operates an algorithmic model that automatically selects AVSs and operators based on yield and risk parameters set by the protocol team. The hands-off approach reduces user friction but centralizes risk assessment within a small development team. Kelp DAO's TVL recovered to $874 million from a low near $500 million following the April 2026 exploit, but remains well below its pre-exploit levels above $1.2 billion.
On April 18, 2026, attackers drained 116,500 rsETH — approximately $292 million — from Kelp DAO's cross-chain bridge infrastructure, according to CoinDesk and Chainalysis. The exploit was the largest DeFi hack of 2026 and exposed structural vulnerabilities in the liquid restaking stack.
The attack vector was instructive. According to OpenZeppelin's post-incident analysis, no smart contract vulnerability was found in Kelp DAO's code or in the rsETH token logic. Instead, attackers — attributed to North Korea's Lazarus Group by Chainalysis — compromised two RPC nodes feeding data to LayerZero's Decentralized Verifier Network (DVN). A simultaneous DDoS attack knocked legitimate verification channels offline. The attackers then injected a forged cross-chain message, authorizing the drain.
The stolen rsETH was immediately deposited on Aave as collateral to borrow $190 million in WETH — against assets now backed by nothing. Because Kelp DAO's bridge held reserves backing rsETH across more than 20 networks, the loss raised doubts about rsETH backing on every layer-2 where it circulated. Aave, SparkLend, and Fluid froze rsETH markets within hours.
CoinDesk reported that Kelp DAO subsequently claimed LayerZero had approved the bridge configuration that was exploited. LayerZero disputed this characterization. The incident highlighted that LRT security depends not only on Ethereum-layer slashing conditions but on the entire cross-chain infrastructure stack — bridges, oracles, RPC providers — none of which fall under EigenLayer's slashing framework.
Academic scrutiny of restaking's risk profile has intensified. An IEEE study published in 2026 ("Systemic Risk in Ethereum's Restaking Architecture: An Empirical Study of Liquid Restaking Protocols") found that many AVSs are secured by highly centralized validator sets. The reuse of collateral, combined with centralized delegation and complex smart contract architectures, creates conditions for what the authors term "risk amplification and correlated failures."
A separate paper published on arXiv in April 2026 ("Financial Dynamics and Interconnected Risk of Liquid Restaking") examined how leveraged restaking positions — where users deposit LRTs as collateral on lending protocols, borrow additional ETH, and restake again — create recursive exposure. The same initial capital can underpin multiple layers of positions. If any single layer fails, losses cascade through the entire structure.
Ethereum's co-founder Vitalik Buterin warned as early as 2023 about "consensus overload" — the risk that Ethereum validators, through restaking, take on obligations that extend beyond Ethereum's core consensus. His concern: if a major AVS suffers a catastrophic failure, the financial fallout for Ethereum validators could be large enough to threaten Ethereum's own security model.
ESMA's September 11, 2026 risk report flagged deepening linkages between crypto and traditional finance as a systemic concern. While not specifically addressing restaking, the regulator noted that DeFi exploits and rising institutional participation create transmission channels through which crypto-native shocks could reach broader financial markets. Crypto market capitalization fell to approximately €1.9 trillion by end-June 2026, a 28% year-to-date decline.
The concentration numbers compound these concerns. EigenLayer holds 93.9% of restaking market share. Ether.fi holds 61% of LRT market share. A failure at either protocol level would have outsized effects across the Ethereum ecosystem.
The EIGEN token presents a stark disconnect between protocol adoption and token performance. EigenCloud's TVL stands at $19.7 billion. EIGEN trades at approximately $0.21, according to CoinGecko and Metamask data, down 96% from its all-time high of $5.65 set in late 2024. Market capitalization sits at approximately $184 million.
The divergence reflects several factors. Monthly token unlocks from vesting schedules create persistent sell pressure — the September 1, 2026 unlock alone released 39.5 million tokens. Bitfinex's delisting removed a source of trading liquidity. More fundamentally, EIGEN's tokenomics do not directly capture restaking revenue. Fees flow to operators and stakers; EIGEN functions primarily as a governance token with future utility promises around intersubjective slashing — a mechanism still in development.
This pattern is consistent with broader DeFi token dynamics observed across protocols where usage metrics diverge from token valuations. The gap between $19.7 billion in TVL and $184 million in market cap suggests the market does not currently price EIGEN as a claim on restaking cash flows.
The restaking thesis — that staked ETH can simultaneously secure Ethereum and dozens of auxiliary protocols — has achieved scale. Five million ETH and $19.7 billion in TVL are not trivial numbers. The EigenCloud rebrand signals ambitions beyond shared security toward a full verifiable cloud infrastructure.
The risk profile has also scaled. The Kelp DAO exploit showed that LRT contagion channels extend far beyond what on-chain slashing mechanisms can govern. Academic research confirms that validator centralization and leveraged positions create correlated failure conditions. The EIGEN token's 96% drawdown suggests the market has already priced in meaningful uncertainty about how — or whether — protocol growth will translate into token value.
Restaking is no longer an experiment. It is a $19.7 billion infrastructure layer embedded in Ethereum's security architecture. Whether the economic model sustains itself under real slashing conditions, repeated exploits, and compressed yields remains an open question. The data suggests the answer will arrive through stress events rather than theoretical analysis.