ZK’s New Bottleneck: Why Aligned Is Retiring Its Verification Layer for Aggregation

On 21 July, Aligned announced that it is deprecating its Proof Verification Layer (PVL) and making its Proof Aggregation Service the default and only solution going forward. The company says new tasks are no longer being sent to PVL, while operators have been asked to deregister and shut down their running instances.

This is not the launch of a new proof system. It is a more revealing infrastructure decision: a ZK provider is concluding that a standalone verification layer is less compelling than aggregation tied to Ethereum’s security model.

Aligned’s original premise was that Ethereum verification costs, throughput, and latency warranted a dedicated operator network. Its stated rationale for the pivot is threefold: proving systems have improved; Ethereum is moving toward more native ZK capabilities; and customers have preferred Ethereum-level security guarantees over the lowest possible latency. In other words, the company is betting that the value is moving from operating another verification layer to making proofs fit more directly into Ethereum’s settlement and security environment.

That distinction matters for builders. “ZK infrastructure” is not a single product category. A prover creates a proof, a verifier checks it, and an aggregation service can package the verification path around a shared security model. Aligned’s change does not make those functions disappear; it changes which one it intends to operate as its focal product. Teams using the retiring layer should therefore assess migration, service dependencies, and the security assumptions of the aggregation path—not assume that every ZK workload has become interchangeable.

The broader Ethereum roadmap helps explain why this is timely. Ethereum.org describes L1 zkEVM verification as a move from every validator re-executing a block toward a model in which a prover executes it and validators verify a proof. That could eventually ease validator hardware pressure, but it also creates a demanding proving problem. The same roadmap notes that current zkEVM implementations can take minutes to hours to prove a block, while useful consensus integration would need proofs on a much tighter schedule.

So Aligned’s retreat from a separate verification layer should not be read as proof that Ethereum-native ZK has arrived. It is better understood as a market participant simplifying its architecture around the path it sees as durable: aggregation backed by Ethereum’s existing security rather than another intermediate network.

Caveat: Aligned’s reported 136,000-proof count, 59-operator figure, and cost/throughput claims are self-reported and were not independently audited for this article. Separately, Ethereum’s L1 zkEVM verification remains active research and is not integrated into production Ethereum clients.

The useful takeaway is narrower but important: as proving improves, ZK infrastructure may compete less on adding layers and more on how credibly it connects proofs to the chain that ultimately secures them.