# Aztec Discloses a Critical Soundness Flaw in Its Alpha V5 Proving System

Core contributors found the defect through internal AI-assisted auditing and say a crafted proof could pass verification for a transaction the network should reject, with the fix deferred to V6.

- Published: 2026-09-28T05:43:34.972Z
- Canonical: https://polylog.news/crypto/2026-09-28/aztec-discloses-a-critical-soundness-flaw-in-its-alpha-v5-pr
- Publisher: Polylog (Crypto desk)
- Section: crypto
- Sources: [Aztec Network](https://aztec.network/blog/alpha-v5-proving-system-vulnerability), [Aztec Network](https://aztec.network/alpha-v5), [Miden](https://miden.xyz/prifi-stack)

Aztec, the privacy-focused layer-2 network on Ethereum, published a disclosure describing a [critical vulnerability in the proving system behind its Alpha V5 network](https://aztec.network/blog/alpha-v5-proving-system-vulnerability). Core contributors identified it on 27 July 2026 through internal auditing assisted by artificial intelligence models. The defect would allow an attacker to construct a proof that passes verification for a transaction the protocol should reject, which puts V5 funds, applications and contract state at risk. The team says the repair ships with V6.

This is a soundness failure, not a smart-contract bug. On a network where execution is private and correctness is asserted by a proof, there is no set of validators re-running the transaction to notice that it was invalid. That distinction is why proving-system defects have to be handled with embargoed disclosure timed to an upgrade, and it is the second such finding for the network after a [critical flaw in Alpha v4](https://aztec.network/alpha-v5) was disclosed earlier in the year.

The disclosure comes as privacy-focused networks seek institutional adoption. Miden, another zero-knowledge network, is marketing a [stack for confidential but verifiable and compliant finance](https://miden.xyz/prifi-stack) with client-side proof generation and local account state. Both projects argue that confidentiality is a prerequisite for serious on-chain settlement, which makes the reliability of their proof systems central to that argument.

## What this means

The exposed parties are users of privacy rollups still in testing and, indirectly, every team selling confidential execution to regulated institutions, because a soundness bug in a proof system breaks the one guarantee that replaces public verification. Attackers gain an advantage invisible on-chain, and auditors with formal-methods expertise gain work. The framing also benefits Aztec directly: it presents the flaw as evidence that its alpha process works as intended, a claim that is credible and also self-serving. The precedent is the Merge-era pattern of embargoed client bugs patched before mainnet, where early disclosure limited the damage. The likelier outcome over the next several months is that privacy networks compete on verification assurances, including multiple proving backends, published audits and reliable withdrawal mechanisms, before they compete on fees. That expectation weakens if a proving-system flaw is exploited on a live network holding real deposits, in which case institutional interest in confidential execution slows instead of growing.

## What to watch

- Whether the V6 release ships the fix with an independent audit attached, the test of whether Aztec's disclosure process converts into verified code.
- Whether other zero-knowledge networks adopt multiple proving backends, which would show the industry now treats soundness bugs as a standing risk class.
- How many privacy-network flaws are first surfaced by AI-assisted review, since that would shift the balance between attacker and defender in cryptographic code.