Claude AI Exposes Critical Zcash Vulnerability

Independent security researcher Taylor Hornby discovered a critical soundness vulnerability in Zcash's Orchard shielded-pool circuit on May 29, 2026, while auditing the protocol for Shielded Labs, using Anthropic's Claude Opus 4.8 and a custom AI tool to write a working exploit that minted counterfeit ZEC in a local test environment. The flaw, an under-constrained element in the zero-knowledge proof circuit present since May 2022, allowed mathematically invalid inputs to pass an elliptic-curve check, prompting an emergency soft fork on June 2 and a hard fork on June 3 to patch the bug. The Zcash Foundation reported no evidence of exploitation or unauthorized value creation, though ZEC fell roughly 37% intraday before recovering.

Claude AI Exposes Critical Zcash Vulnerability Multiple outlets, including The Block and Gizmodo, report that independent security researcher Taylor Hornby discovered a critical soundness bug in Zcash's Orchard shielded-pool circuit on May 29, 2026, while auditing the protocol for Shielded Labs. Per these reports, Hornby used Anthropic's Claude Opus 4.8, alongside a custom AI tool, to write a working exploit that minted counterfeit ZEC in a local test environment. The flaw, an under-constrained element in the zero-knowledge proof circuit, let mathematically invalid inputs pass an elliptic-curve check. After private disclosure, developers shipped an emergency soft fork on June 2 disabling Orchard, then the NU6.2 hard fork on June 3 re-enabled it with a corrected circuit. The Zcash Foundation says there is no evidence of exploitation or unauthorized value creation. ZEC fell roughly 37% intraday before recovering much of the loss. What happened Independent security researcher Taylor Hornby discovered a critical soundness vulnerability in the Orchard shielded-pool circuit of Zcash on May 29, 2026, while conducting a protocol audit for Shielded Labs, according to The Block, Gizmodo, and the Zcash Foundation. Reports state Hornby used Anthropic's Claude Opus 4.8, together with a custom AI tool, to write a working exploit that generated counterfeit ZEC in a local test environment. The discovery set off an emergency, coordinated response across the Zcash ecosystem and a sharp move in the token's price. The vulnerability Per The Block and the Zcash Foundation, the flaw was an under-constrained element in the Orchard zero-knowledge proof circuit in the halo2 gadgets code , present since the Orchard pool launched in May 2022. The bug let mathematically invalid inputs pass an elliptic-curve check that should have rejected them, which could have allowed an attacker to forge counterfeit ZEC or push invalid shielded transactions through verification. The Zcash Foundation framed the risk carefully, stating there was no evidence the bug was exploited, no unauthorized value creation, and no impact on user privacy. The emergency response After private disclosure to ecosystem engineers and coordination with miners and exchanges, developers shipped a two-phase fix, according to the Zcash Foundation and CryptoBriefing. An emergency soft fork activated at mainnet block 3,363,426 on June 2, temporarily disabling Orchard transactions while corrective code was prepared. The NU6.2 hard fork then activated at block 3,364,600 on June 3, re-enabling Orchard with a corrected circuit. Reports note NU6.2 required a hard fork because fixing a zero-knowledge circuit bug means updating the pinned verifying key, which a routine node patch cannot change. The Foundation called it only the second security-driven protocol upgrade in the network's history since 2016. Market reaction ZEC fell sharply on the news, with The Block citing a drop of about 31% and other outlets citing roughly 37% over 24 hours, touching an intraday low near $385 before recovering much of the loss as the patched upgrade activated. The divergent figures reflect different measurement windows rather than a settled number. Editorial analysis LLM-assisted source-code review is an emerging pattern in which large models help surface edge-case logic and boundary-condition errors that human reviewers can miss. This incident is a concrete, high-stakes example: a frontier model was used not only to spot a subtle soundness flaw in production cryptographic code but to write a functioning proof-of-concept exploit. For security teams, it suggests AI tools can meaningfully accelerate triage and hypothesis generation in complex, math-heavy codebases. The episode also reinforces that model output requires verification against formal proofs, reproducible test vectors, and conservative change control before consensus-layer code is trusted. What to watch - •Whether independent auditors publish technical write-ups or formal-verification traces that reproduce Hornby's findings. - •Whether on-chain monitoring or supply-proof proposals from groups such as Shielded Labs provide ongoing assurance against undetected minting. - •Whether more protocol teams adopt LLM-assisted auditing as a standard part of pre-deployment review. Scoring Rationale A frontier LLM Claude Opus 4.8 was used to discover and build a working exploit for a critical soundness bug in Zcash's core Orchard circuit, forcing an emergency soft fork and the NU6.2 hard fork, only the second security-driven upgrade in the network's history. Broad independent coverage and the landmark AI-assisted-security angle make it highly relevant to AI and security practitioners. It stops short of top scores because no supply was actually inflated and the issue was patched with no evidence of exploitation. Practice with real FinTech & Trading data 90 SQL & Python problems · 15 industry datasets Active Verified Users by Income TierEasy /problems/sql/active-verified-users-by-income Technology Stocks with High BetaMedium /problems/sql/technology-stocks-with-high-beta Portfolio Performance ScorecardHard /problems/sql/portfolio-performance-scorecard 250 free problems · No credit card See all FinTech & Trading problems /problems/datasets/fintech