Nearly $40 billion in total value locked sits fragmented across more than 20 competing Ethereum Layer 2 networks. On March 29, 2026, at EthCC in Cannes, Gnosis co-founder Friederike Ernst and Zisk founder Jordi Baylina unveiled the Ethereum Economic Zone (EEZ) — a rollup framework co-funded by th...
"Ethereum doesn't have a scaling problem. It has a fragmentation problem. Every new L2 that launches with its own liquidity pool and its own bridge is another walled garden." — Friederike Ernst, Co-founder, Gnosis
Nearly $40 billion in total value locked sits fragmented across more than 20 competing Ethereum Layer 2 networks. On March 29, 2026, at EthCC in Cannes, Gnosis co-founder Friederike Ernst and Zisk founder Jordi Baylina unveiled the Ethereum Economic Zone (EEZ) — a rollup framework co-funded by the Ethereum Foundation that promises synchronous cross-rollup composability without bridges. The EEZ enters a crowded field: Optimism's Superchain (34 chains, ~$20.4B TVL), Polygon's AggLayer, and the Ethereum Foundation's own Interop Layer (announced November 2025) are all pursuing the same goal.
The stakes are structural. While L2s collectively reduced Ethereum transaction costs by orders of magnitude, they fractured liquidity, broke DeFi composability, and created user experiences that require manual bridging between siloed networks. Ethereum mainnet retains $106.57 billion in TVL but captured less than $10 million in L2 settlement fees in 2025 — a 90%+ year-over-year decline. Four competing interoperability frameworks now race to reunify the ecosystem, each with different technical architectures and different trade-offs.
Data from L2BEAT shows the following market share distribution among major L2 networks as of late March 2026:
| Network | Category | Approx. TVL | DeFi TVL Share | |---------|----------|-------------|----------------| | Arbitrum One | Optimistic | ~$17-19B | ~31-35% | | Base | Optimistic | ~$4-15B | ~33-47% | | OP Mainnet | Optimistic | ~$3-8B | ~6% | | zkSync | ZK-Rollup | Varies | <5% | | Scroll | ZK-Rollup | Varies | <3% | | Starknet | ZK-Rollup | Varies | <3% | | 14+ others | Mixed | Varies | <10% combined |
Note: TVL figures vary by source and measurement methodology. DeFi-specific TVL differs from broader total value secured metrics. Figures cited draw from DefiLlama, L2BEAT, and The Block's 2026 Layer 2 Outlook.
The numbers illustrate a structural problem. Arbitrum and Base together represent over 75% of L2 DeFi TVL, according to The Block's 2026 Digital Assets Outlook. The remaining value is scattered across more than 18 smaller networks — each with its own bridge, sequencer, and liquidity pool.
Consequences are measurable. According to multiple sources, liquidity fragmentation reduced average pool depth by approximately 40% across L2 networks. Users face fees on both sides of bridge transactions, wait for confirmations, and navigate interfaces that vary by network. Bridge-related security failures accounted for approximately $2.8 billion in losses in 2025, representing roughly 40% of all Web3 exploits that year.
For developers, the fragmentation forces a deployment decision: which L2 to target. A lending protocol deployed on Arbitrum cannot natively access liquidity deposited on Base. A user holding ETH on Optimism cannot purchase an NFT on Scroll without first bridging — a process involving fees, latency, and counterparty risk.
The L2 interoperability landscape has consolidated around four distinct approaches, each backed by different stakeholders with different technical philosophies.
The EEZ, announced at EthCC Cannes on March 29, makes a specific technical claim: synchronous composability between Ethereum mainnet and connected rollups within a single transaction. This means a smart contract on one EEZ-connected rollup could call a contract on mainnet — or on another connected rollup — with the same execution guarantees as a same-chain call.
The enabling technology, according to Zisk founder Jordi Baylina, is a zero-knowledge virtual machine (ZKVM) that can prove Ethereum blocks in real time. Baylina, who created the Circom programming language and co-founded Polygon zkEVM before spinning off an independent team into Zisk in June 2025, stated at EthCC: "We spent two years building a ZKVM that can prove Ethereum blocks in real time."
Founding members include DeFi lending protocol Aave, block builders Titan and Beaver Build, real-world asset platform Centrifuge, and tokenized equities project xStocks. The project will be structured as a Swiss non-profit with all software released as free and open-source.
GnosisDAO governance records from February 2026 show the community had already been debating a six-month R&D collaboration with Baylina to explore converting Gnosis Chain into a natively integrated Ethereum L2 with synchronous composability. The EEZ appears to be the product of that exploration.
A Gnosis spokesperson indicated the first testnet should be ready by mid-2026. Technical benchmarks for the real-time proving stack have been promised but not yet published.
Assessment: The EEZ's real-time ZK proving claim, if validated, would represent a meaningful technical advance. However, Zisk's proving stack capabilities remain unverified by independent parties. The framework also requires voluntary adoption from existing L2 teams — none of the major rollups (Arbitrum, Base, OP Mainnet) have indicated intent to join.
Optimism's Superchain represents the most established interoperability framework by deployment count. It comprises 34 OP Stack-based chains, according to Optimism documentation, and accounts for over 50% of all L2 activity and more than 10% of all crypto transaction volume.
The Superchain's interoperability model relies on shared infrastructure: a common bridge, shared sequencer plans, and the OP Stack as a standardized deployment framework. The full Interop Layer rollout, planned for early 2026, introduces native message passing, ERC-7802-compliant asset bridging, and shared fault proofs.
The SuperchainERC20 token standard enables cross-chain portability and fungibility across all OP Stack chains. With interoperability on mainnet, data and assets would move between chains without transiting through L1, a requirement that currently adds latency and cost.
Base — Coinbase's L2, built on the OP Stack — is the Superchain's largest venue, with approximately $4.09 billion in TVL as of March 22, 2026, representing roughly 65.9% of total Superchain TVL (per Messari's H1 2025 State of the Superchain report methodology).
Limitation: The Superchain is strictly limited to OP Stack chains. It does not extend interoperability to ZK-rollups or non-OP Stack deployments. This creates a walled garden of its own — unified internally, but isolated from Arbitrum, zkSync, Starknet, and others.
Polygon's AggLayer takes a different architectural approach. Rather than requiring chains to share a common stack, AggLayer uses a unified bridge and pessimistic proofs to coordinate across chains built with Polygon's Chain Development Kit (CDK).
The Unified Bridge eliminates token wrapping and unwrapping, enabling native asset transfers across all connected chains through a shared pool. A new chain connecting to AggLayer automatically accesses the shared asset pool of all participating networks — a design that reduces the cold-start liquidity problem for new deployments.
By mid-2026, the CDK Enterprise edition enabled institutions to deploy permissioned EVM chains with financial-grade privacy, while maintaining connection to the broader AggLayer liquidity pool.
Limitation: Adoption remains narrow. The AggLayer currently connects a small number of chains, with Polygon zkEVM as the primary venue. Compared to the Superchain's 34 chains, AggLayer's network effects are limited.
The Ethereum Foundation's Account Abstraction team published the Ethereum Interop Layer (EIL) specification in November 2025, describing a system that would make Ethereum's rollups "feel like one chain" while preserving trust-minimized, decentralized foundations.
The EIL is built on ERC-4337 account abstraction. Users sign once for a cross-chain transaction without adding new trust assumptions. The design is wallet-native: Ethereum wallets and dapps become "multichain-native by default," and new and existing rollups are "automatically compatible."
Unlike the Superchain or AggLayer, the EIL does not require chains to use a specific stack or connect to a specific bridge. It operates at the account layer, making it theoretically compatible with any L2 that supports ERC-4337.
Limitation: The EIL is the least mature of the four frameworks. Future updates promise faster confirmations and real-time proofs, but production deployment timelines remain unspecified. Its account-abstraction-only approach may not satisfy applications requiring atomic, synchronous composability.
The interoperability question is ultimately an economic question: who captures the value generated by cross-chain activity?
Under current architecture, L2s retain the vast majority of transaction fee revenue. Coinbase's Base earned over $94 million in profit in 2025 but contributed just $4.9 million to Ethereum mainnet in blob fees. OP Mainnet retained $321.31 for every $1 paid to mainnet security. Arbitrum One's ratio was $28.62 to $1.
Total fees paid by L2s to Ethereum mainnet collapsed to roughly $10 million in 2025 — a 90%+ year-over-year decline, representing less than 10% of L2 revenue. The introduction of EIP-4844 (Proto-Danksharding) in March 2024 made L2 data availability so cheap that blob fees frequently fell to zero.
Each interoperability framework implies a different value capture model:
| Feature | EEZ | Superchain | AggLayer | EIL | |---------|-----|------------|----------|-----| | Architecture | Sync ZK composability | Shared OP Stack | Unified bridge + CDK | Account abstraction | | Proof System | Real-time ZK (claimed) | Fault proofs | Pessimistic proofs | N/A (wallet layer) | | Chain Requirement | EEZ-compatible rollup | OP Stack only | Polygon CDK | Any ERC-4337 chain | | Connected Chains | 0 (pre-testnet) | 34 | ~3 | Open specification | | Cross-chain Latency | Single transaction (claimed) | Sub-minute (planned) | Bridge-dependent | Signing-dependent | | Governance | Swiss non-profit | Optimism Collective | Polygon Labs | Ethereum Foundation | | Mainnet Settlement | Direct to Ethereum L1 | Via shared bridge | Via Unified Bridge | Via underlying L2 | | Maturity | Announced March 2026 | Operational since 2024 | Launched Q2 2025 | Spec published Nov 2025 |
$40 billion in L2 TVL is split across 20+ networks, with Arbitrum and Base capturing over 75% of DeFi TVL. The remaining 25% is scattered across 18+ smaller chains, each operating as a liquidity silo.
Four competing interoperability frameworks have emerged in the span of 18 months. None has achieved cross-framework compatibility. The risk of interoperability itself becoming fragmented — four incompatible unification layers — is non-trivial.
The EEZ's real-time ZK proving claim is technically ambitious but unverified. No independent benchmarks have been published. The testnet is projected for mid-2026. Founding members (Aave, Centrifuge, Titan) add credibility but do not validate the underlying cryptography.
Superchain holds the adoption lead with 34 operational chains and over 50% of L2 transaction activity. Its limitation is exclusivity: OP Stack chains only.
Value capture remains asymmetric. L2s retained over 90% of fee revenue in 2025, with mainnet capturing less than $10 million total. Whichever interoperability framework wins may determine whether Ethereum L1 regains economic relevance in its own scaling ecosystem.
Bridge security remains an open risk. Approximately $2.8 billion was lost to bridge exploits in 2025. Any interoperability framework that introduces new bridge mechanisms inherits this attack surface.
Ethereum's Layer 2 ecosystem faces a paradox: scaling succeeded, but unification failed. The network now processes more transactions at lower cost than at any point in its history, yet users and developers navigate a fragmented landscape that undermines the composability that made Ethereum valuable in the first place.
The EEZ announcement adds a fourth contender to an interoperability race that already includes three well-funded frameworks. Whether four competing unification layers produce convergence or further fragmentation remains to be seen. The Ethereum Foundation's decision to co-fund the EEZ while simultaneously developing its own Interop Layer suggests the Foundation itself does not yet view any single approach as sufficient.
The market will ultimately select winners based on adoption, not architecture. As of March 30, 2026, the Superchain leads on deployment count, the AggLayer leads on institutional integration, the EIL leads on theoretical compatibility, and the EEZ leads on technical ambition. None leads on all four dimensions simultaneously.
For Ethereum mainnet, the economic question looms. If L2s continue to capture 90%+ of fee revenue while settling to L1 for less than $10 million annually, the interoperability layer that eventually wins will determine not just user experience — but whether Ethereum's base layer retains economic relevance in the ecosystem it created.