Advanced AI models could still uncover bugs in Aztec Tierra (Alpha V6), the latest version of the Ethereum privacy network, while only its most security-critical code has been formally verified, Aztec Foundation co-founder and CEO Zac Williamson said.
Tierra launched Nov. 3. In an interview recorded Nov. 9, Williamson said he “can’t rule out the possibility that bugs will be found by more advanced AI models.” He said Aztec’s AI audits of the previous version found critical bugs in cryptographic libraries, and that the team runs audits around the clock on its own testing system.
A flaw in V5’s proving system could have allowed a forged proof to pass verification. It remains unclear whether the flaw was exploited. Tierra fixes the V5 flaw and launched with an AI testing harness and formal verification for core cryptographic components.
Williamson said Aztec has not formally verified the entire protocol. After AI-assisted audits found critical bugs, the team formally verified the affected subsystems. Formal verification uses mathematical proofs to check whether code behaves as intended, but the work so far covers only the most security-critical parts of Aztec’s protocol.
For major network upgrades, Williamson said Aztec effectively uses complete hard forks: users withdraw from the old version and deposit into the new one. He said withdrawals of tokens held on Ethereum would expose a hidden double-spend in the old version, and that the issue “can’t extend to the new version of Aztec.”
Williamson also said Aztec does not have a security council like other networks. He described formal verification of the full protocol as the long-term solution.
Simon Yoon
Simon Yoon reports on blockchain technology for TokenPost. Send corrections or tips to [email protected].


