Is L2 really secured by Ethereum?

Chainfeeds Guide:

Almost all major Rollup projects, including Arbitrum, Optimism, Base, zkSync, and Scroll, use “secured by Ethereum” as the core of their brand. This slogan is powerful and forms the heart of their marketing narrative, but does it really reflect reality?

Source:

Author:

Hazeflow

Opinion:

Hazeflow: Over the past decade, Ethereum has been pursuing a simple promise: to scale the network without sacrificing decentralization. The solution is the “Rollup-centric” roadmap, allowing L2s (Layer 2s, commonly referred to as rollups) to execute transactions off-chain for lower costs and higher throughput, while relying on Ethereum itself for core security. But when almost all mainstream Rollups—Arbitrum, Optimism, Base, zkSync, Scroll—are promoting the slogan “secured by Ethereum,” the reality is far more complex than the slogan suggests. What users actually interact with is not the internal math of the Rollup, but the bridge. The bridge is the channel for assets to enter and exit the Rollup, determining whether users can safely retrieve their funds. Depositing is relatively simple: you deposit 1 ETH into a bridge contract on Ethereum, it gets locked, and your account on the Rollup shows 1 ETH—this part almost entirely relies on L1 security. But withdrawals are much more complicated: because Ethereum cannot directly see what happens inside L2, it can only rely on proofs provided by the bridge. There are roughly three types of proof models: first, the optimistic (fraud proof) model, which assumes transactions are valid unless someone challenges them during a dispute period; second, the zero-knowledge (validity proof) model, which uses cryptography to ensure every transaction is compliant, allowing Ethereum to immediately trust the result; third, multisig or committee endorsement, relying on a group of trusted external entities. In other words, different bridge models carry vastly different risks. Users often think their funds are protected by Ethereum, but in reality, Ethereum only secures the locked portion in the “official bridge”; other bridges, especially third-party bridges, rely more on external operators. The real conclusion is: no matter how decentralized the Rollup itself is, if the bridge is compromised, users cannot safely move funds in or out. The bridge is the most fragile yet crucial “window” in the Rollup system. As of August 29, 2025, Rollups in the Ethereum ecosystem collectively hold about $43.96 billion in assets. But the distribution is highly uneven: $16.95 billion (39%) enters through external bridges, $14.81 billion (34%) is locked in official bridges, and the remaining $12.2 billion (27%) are assets natively issued within the Rollups. From 2019 to 2022, almost all growth relied on official bridges, with Ethereum’s security still at the center. But since the end of 2023, the landscape has changed significantly: on one hand, the total amount in official bridges is still rising but gradually losing its share; on the other hand, native asset issuance is steadily expanding, especially gaining scale in 2024–2025; the most dramatic change comes from external bridges, which, thanks to faster user experience (especially for cross-chain transfers and quick withdrawals), accelerated sharply in 2024, even surpassing official bridges in early 2025, causing Ethereum’s direct security coverage of Rollup assets to fall below half. Now, two-thirds of Rollup assets are outside Ethereum consensus. Looking at the top six Rollups, they account for as much as 93.3%. Their internal structures also show different tendencies: Arbitrum and Unichain lean more toward external bridges, with users seeking liquidity and speed; Linea and some Optimism instances prefer official bridges, emphasizing orthodox L1 guarantees; zkSync Era and Base are dominated by native issuance, such as native USDC on Base. This means that while everyone is shouting “secured by Ethereum” on the surface, in reality, most assets have already drifted beyond Ethereum’s direct security boundary—the bridge model and issuance logic are the real dividing lines. Even if all assets enter through official bridges, users still face three types of structural risks: sequencing, governance, and composability. Currently, almost all Rollups rely on a centralized sequencer to order transactions—it decides the order in which transactions are batched. The benefit is speed, but the downside is obvious: it can censor transactions and refuse to include them; it can also delay withdrawals or even block users from withdrawing altogether; more likely, an outage could bring the entire network to a halt. Although Ethereum has designed a “force inclusion” mechanism, allowing users to submit transactions directly on L1, this is only an inefficient safety valve, not a guarantee of fairness. The sequencer can still front-run your transaction, changing the outcome. For example, you initiate a withdrawal on L2, but the sequencer inserts a lending operation before yours, draining the liquidity pool—your withdrawal is included but fails. On the other hand, governance risk is also prominent. Arbitrum, Optimism, Base, etc., are operated by companies or VC teams, whose shareholder obligations take precedence over Ethereum community contracts. This leads to a “platform tax” path: low fees attract users in the early stage, but fees are raised after locking them in; policy-making and upgrades are mostly decided by a few management keys, leaning more toward business compliance than decentralized governance. Over time, Rollups may become more like closed corporate gardens rather than open extensions of Ethereum. External bridges, centralized sequencers, and corporate governance—these three elements together are widening the gap between Ethereum’s ideals and real-world implementation. In the future, shared sequencers (such as Espres