Aztec Says a Flaw in Its Proving System Puts Funds on Its Live Privacy Network at Risk

Aztec, the privacy-focused layer-2 network on Ethereum, has disclosed a critical vulnerability in the proving system behind Alpha V5, the version of its network that governance activated after the upgrade cut the cost of proving a private transaction in half.

The defect is a soundness failure, not a leak of private data. According to the disclosure, an attacker may be able to construct a proof that passes verification for a transaction the network should reject. If the network accepted such a proof, the resulting state transition would fall outside the rules V5 is supposed to enforce. Contributors say they cannot determine whether anyone exploited the flaw before they found it. They advise treating funds, applications and contract state on Alpha V5 as exposed to a protocol-level failure until incident response finishes and node operators complete the required network actions.

Two details matter more than the bug itself. First, contributors say they found it through internal auditing assisted by artificial-intelligence tooling, the same class of review that is now surfacing defect backlogs across open-source cryptography. Second, they state that the flaw is not mitigated by public re-execution by the committee of validators. Re-execution is the fallback most rollups rely on when a prover misbehaves, because honest nodes can recompute a disputed block. A proof that verifies for an invalid statement defeats that check, because there is nothing for validators to disagree about at the level they inspect.

This is the second proving-system emergency in Aztec's alpha cycle. Contributors disclosed a critical vulnerability in Alpha V4 earlier this year and told users to withdraw before the V5 governance vote made the V4 details public. V5 was the fix. The network is explicitly labelled alpha software, and a V6 release is planned for later this year alongside a published security roadmap, but the ecosystem around it is already live, including applications and games running on the network.

The core problem is that privacy networks carry a compounding verification burden. A transparent rollup can be audited by anyone who replays the chain. A private one cannot, so the proof system is not one safeguard among several, it is the safeguard. That raises the cost of shipping and lengthens the path from alpha to a network that custodians and institutions will hold balances on.