According to blockchain security analysts, AI technology is enabling cybercriminals to quickly identify weaknesses in smart contracts, resulting in $1.32 billion being stolen in the first half of 2026, with outdated protocols increasingly becoming the focus of these attacks.
Latest in Crypto Security
Experts in blockchain security assert that a single audit of smart contracts is no longer sufficient for ensuring the safety of decentralized systems. They emphasize that AI tools are now helping malicious actors uncover vulnerabilities much faster than traditional auditing methods can manage.
Related Article: ETH Trader Faces $2M Loss Due to Same-Block DEX Backrun Attack
The Rise of Old Code as a Target
On June 14, a smart contract vulnerability was exploited to siphon $2.1 million from Aztec Connect, a protocol that ceased operations in March 2023. Despite the shutdown, the smart contract remained active on-chain, offering attackers an unsecured access point years after the service ended. Just five days later, $300,000 was stolen from a smart contract associated with the decentralized exchange mySwap, which had stopped taking new liquidity deposits over six months prior to the attack.
The Zcash (ZEC) network demonstrates the dual-edged role of AI in cybersecurity. Taylor Hornby, a security engineer at Shielded Labs, utilized a custom auditing tool based on Anthropic’s Claude Opus 4.8 to detect a significant flaw in Zcash’s Orchard shielded pool—a crucial privacy component. This vulnerability had remained unseen for four years and could have allowed illicit counterfeiting within the pool before it was addressed following Hornby’s discovery.
The Necessity of Ongoing Review
Ari Redbord, the policy head at TRM Labs, shared with Cointelegraph that ongoing review is proving more effective than one-time audits. “Attack strategies evolve more rapidly than what a singular audit can address from launch day,” he pointed out. According to him, an audit reflecting last year’s threats fails to protect against current methods employed by attackers.
CertiK has advised projects operating on outdated infrastructure to view re-audits as an ongoing expense rather than a one-off task upon deployment. Redbord further indicated that while robust code is essential, it is only part of the solution; broader challenges, such as countering North Korean cyberattacks and Chinese laundering networks that channel stolen funds, must also be tackled. “Protocols may secure their entrances,” he emphasized, “but it is crucial to pursue the actors attempting to breach them.”
This article features links to external websites or content for informative purposes only (“Third-Party Sites”). CoinMarketCap does not control these sites and is not responsible for their content or any changes to them. Links are provided solely for convenience, and their inclusion does not indicate endorsement or approval from CoinMarketCap. This content is meant for informational use only. It’s essential to conduct thorough research before making significant decisions related to any products or services mentioned. The views expressed in this article are solely those of the author or company and do not necessarily reflect the opinions of CoinMarketCap.