Zcash Audit Finds No New Critical Bugs After Anthropic’s Mythos AI Security Review

A security review of the Zcash network using Anthropic’s Mythos artificial intelligence model has found no additional serious vulnerabilities, according to Zcash founder Zooko Wilcox. The audit was requested by Shielded Labs, a Swiss-based nonprofit that supports the development of the privacy focused cryptocurrency.

The review comes shortly after Zcash developers fixed a significant vulnerability in the Orchard shielded pool. Earlier this month, Orchard transactions were temporarily suspended after security researcher Taylor Hornby discovered a four-year-old forgery bug with assistance from Anthropic’s Claude Opus 4.8 AI model.

Orchard Vulnerability Was Quickly Resolved

Developers released an emergency upgrade and restored Orchard functionality on the same day the issue was identified. The Zcash Foundation said there was no evidence that the vulnerability had been exploited. It also confirmed that no unauthorized value was created and user privacy remained fully protected.

AI Security Tools Create New Challenges

The findings arrive as AI becomes increasingly important in cybersecurity. Anthropic recently stated that its Mythos models identified more than 10,000 high- and critical-severity vulnerabilities in important software systems.

However, concerns remain about the risks of advanced AI. Anthropic suspended access to its Fable 5 and Mythos 5 models after receiving a U.S. government export-control directive linked to national security concerns.

The growing use of AI in security research highlights both its potential to strengthen blockchain networks and the challenges it creates for the wider crypto industry.